134 97 31MB
English Pages 559 [564] Year 1999
Digitized by the Internet Archive in 2022 with funding from Kahle/Austin Foundation
https://archive.org/details/airshiptechnologoOO0Ounse
This book provides a unique and indispensable guide to modern airship design and operation. Airships today incorporate advanced technology including composite materials, complex electronic systems and fly-by-light controls. They demand the latest
theories in aerodynamics, stability and control and require the use of advanced design tools such as numerical finite element structural analysis and computer aided design. This comprehensive and fully illustrated account brings together airship specialists from both universities and industry. After a general introduction, the essentials of aerostatics, aerodynamics, stability and control, propulsion, materials and structures are covered. The following chapters consider weight estimates and control, ground handling and mooring, systems, performance and piloting. The final chapters examine suggestions for improving airship performance, survey unconventional designs, synthesise various design elements, and look at airship roles and economic considerations vital for the success of the airship in the market place. Detailed references are also included. This book will be of interest to airship designers and engineers, professional and student aeronautical engineers, and also to airship enthusiasts.
SEP 0 2 2005
ne cope & ne Acie
teat] ay. te
enw,
told “idan
ai?
hist
darwse? s
quioly, basee Aarne
+ 6
twa
Hy
iil
ras x a4
x ¢
'
bie
je
bd
baw rob ot
|
7
7 =)
— i
Ln) a8
™yj
ps0
am ipwiinil a are Gir ; 7
Pats
stennita
T5D4.
aie
;
4
'
oe
a
|
~
ost oQubeegeeee S- & : a 7 a a :
gvts
:
;
painiho » Hise
ia Lian
.
Stim
y Wik 8 AY
-
)
;
7
y
_ -
1
iTS!
-
:
‘
ipaweas
er®
:
am
a
—
CAMBRIDGE
AEROSPACE
SERIES:
10
General editors: Michael J. Rycroft, Robert F. Stengel
Airship Technology
CAMBRIDGE
AEROSPACE
SERIES
. J. M. Rolfe and K. J. Staples (ed.): Flight Simulation
. P. Berlin: The Geostationary Applications Satellite . . . . . .
M. J. T. Smith: Aircraft Noise N. X. Vinh: Flight Mechanics of High-Performance Aircraft W.A. Mair and D. L. Birdsall: Aircraft Performance M. J. Abzug and E. E. Larrabee: Airplane Stability and Control M. J. Sidi: Spacecraft Dynamics and Control J.D. Anderson: A History of Aerodynamics
. A. M. Cruise, J. A. DMN — OMAN HBRWNHY
Bowles, C. V. Goodall and T. J. Patrick: Principles of Space
Instrument Design S . G.A. Khoury and J. D. Gillett: Airship Technology
Airship Technology G. A. Khoury Imperial College, London, and The Airship Association
J. D. Gillett Formerly of Brunel University and The Airship Association
= CAMBRIDGE 5
By UNIVERSITY
PRESS
3 1336
06923
6140
PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE The Pitt Building, Trumpington Street, Cambridge, United Kingdom CAMBRIDGE UNIVERSITY PRESS The Edinburgh Building, Cambridge CB2 2RU, UK 40 West 20th Street, New York NY 10011-4211, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia Ruiz de Alarcén 13, 28014 Madrid, Spain Dock House, The Waterfront, Cape Town 8001, South Africa
http://www.cambridge.org © Cambridge University Press 1999 This book is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press.
First published 1999 First paperback edition 2004 Typeset in Times Roman [CRC] A catalogue record for this book is available from the British Library Library of Congress Cataloguing in Publication data Airship technology / G. A. Khoury, J. D. Gillett. p.
cm.— (Cambridge aerospace series; 10)
ISBN 0 521 43074 7 (hardback) 1. Airships — Design and construction. I. Khoury, G. A. (Gabriel Alexander), 1947— II. Gillett, J. D. (John David), 1913-1995. Ill. Series.
TL660.A37 1999 629.133°24—-dce21 98-20441
CIP
ISBN 0 521 43074 7 hardback ISBN 0 521 60753 | paperback
Contents
Preface Acknowledgements
Chapter 1
Introduction (G. A. Khoury and J.D. Gillett)
Chapter 2
Basic Principles (E. Mowforth)
page xi Xu
Introduction Principles of aerostatics The aerostatics of the airship The aerodynamics of the airship Unconventional designs
Chapter 3
Aerodynamics (/. Cheeseman)
Basic Assumptions Drag Dynamic forces Slender body theory An estimation method for overall aerodynamic forces and moments Unsteady aerodynamics Aerodynamic parameter estimation References Chapter 4
Stability and Control (M.V. Cook)
Introduction System of axes and notation The equations of motion The linearised equations of motion Dynamic stability analysis Control and response Automatic flight control Flying and handling qualities Acknowledgements References Symbols Subscripts Stability and control derivatives
Chapter 5
Propulsion (J. Cheeseman)
Introduction The propulsor The prime mover
83 88 96 102 104 104 104 LOS 106 106
107 107 107 129
viii Airship Technology Energy sources
Propulsion, performance and control References Chapter 6
Materials (P. Bradley)
Introduction Desirable properties for airship textile materials Development of textiles for airship use Properties of envelope materials Improved laminates Summary References Chapter 7
Structures (C. Luffman)
Introduction Historical General principles and considerations Principal structural groups References Chapter 8
Aerostatics (J. Craig)
Introduction The atmosphere Contained gas Buoyancy and static lift Summary of airship operations Other factors affecting lift Closed or open Airship balance Derivation of gas mass properties
Terms and abbreviations used Chapter 9
Weight Estimates and Control (J. Craig)
Airship Design Weight Weight
mass properties considerations for weight estimation monitoring and control Build weight control and actual airship weight determination Weights aspects of airship operations
Chapter 10
Systems (N. Mayer)
Introduction Pressure control Fuel Ballast
Contents
Electrical Crew Access and maintenance Emergency Auxiliary thrust
Replenishing References
Chapter 11
Ground Handling and Mooring (D. Howe)
Introduction
1x
285 288 290 292 294 299 296
297 Jy)
Survey of the ground handling problem Ground handling and mooring requirements
a 300
Historical review of ground handling techniques Forces on an airship whilst moored The effect of size on ground handling Future ground handling systems Conclusion
302 309 314 315 319
Chapter 12 _ Piloting (D. Burns) Introduction Effects of controls Static heaviness Effects of wind Weather Vectored thrust Ground handling Take-off Pressure height (altitude) Free ballooning Landing Chapter 13.
Performance (R. Hunt)
321 321 Soe 323 326 327 Jou 333 334 337 338 330 345
Mission considerations
345
Maximum ‘in-flight’ fuel usage
345
Engine selection Power off-take
347 350
Engine drives and propellers Ship weight
351 352
Airship sizing
354
Gas lift
Effect of wind
Chapter 14
Improvements (£. Mowforth)
Introduction Buoyancy control by lifting gas manipulation
352
356
SED 359 359
x Airship Technology Boundary layer control Stern propulsion Lenticular geometry Chapter 15
—_Unconventional Designs (G.A. Khoury)
Introduction Shape Lift Power source Structural configuration Lifting gas Control Payload
References Chapter 16
Solar Power (G.A. Khoury)
Introduction Outline of the ‘Sunship’ Solar radiation Solar-powered flight speed Components of the solar power system References
Chapter 17
Roles and economic considerations (R. Hillsdon)
Introduction The utility of airship types The spectrum of potential roles Role suitability The potential market Airship economics
Conclusion References Chapter 18
Design Synthesis (B.G. Wilson)
Introduction Preliminary comments Baseline design layout Baseline evaluation Design trades and sensitivities Technology trends Design synthesis Index
371 377 380 385 385 385 a Pa 409 417 426 428 434 436
441 441 442 444
448 458 473 475 475 475 483 492 496 500 504 504
507 507 507 514 517 521 528 530
EE)
x1
Preface Revival of serious interest in airships in the UK by both industrialists and academics has taken place in the early 1970s with the formation of the international Airship Association and with the quarterly publication of its Journal Airship. Annual symposia have since been held at the Royal Aeronautical Society and, subsequently, the Airship Association enlarged this activity by holding its International Conventions and Exhibitions in Bedford, England, not far from the two famous airship hangers. Recently, the Airship Association has also set up a dedicated web site on the internet (http://www.airship.demon.co.uk). During this period, the design and development of state-of-the-art non-rigid airships for civil and military applications in the UK owed much to the efforts of Mr Roger Munk to whom this book is dedicated. Whether in design and development, or merely in operation, airship activity has also taken place in many other countries including Australia, Canada, China, Brazil, Germany, Hungary, Italy, Japan, Mexico, New Zealand, Russia, South Africa, and the USA.
Given the increased activity world-wide, there is an urgent need for an up-to-date detailed technical book on this subject. Modern airships employ advanced technologies such as composite materials, numerical finite element structural analysis, computer aided design, modern electronic systems, fly-by-light controls and the latest theories in aerodynamics as well as stability and control. Yet today’s airship designers have only textbooks published in the 1920s and 1930s for reference, namely by Burgess: Airship Design (1927), Blakemore: Pressure Airships Part I (1927), Pagon: Pressure Airships Part IT (1927), and Durand: Aerodynamic Theory (1937). This book, therefore, draws on recent experience by bringing together fourteen specialists in different aspects of airship design and operation. These experts were drawn equally from academia and industry thus providing the appropriate balance of theory and practice. Academics contribute the chapters Basic Principles (Dr Edwin Mowforth), Aerodynamics (Professor Ian Cheeseman), Stability & Control (Mr Michael Cook), Propulsion (Professor Jan Cheeseman), Ground Handling & Mooring (Professor Denis Howe), Improvements (Dr Edwin Mowforth), Unconventional Designs (Dr Gabriel Khoury), and Solar Power (Dr Gabriel Khoury). Industrialists
contribute the chapters Materials (Dr Peter Bradley), Structures (Mr Charles Luffman), Aerostatics (Mr John Craig), Weight Estimates & Control (Mr John Craig), Systems (Mr Norman Mayer), Piloting (Mr David Burns), Performance (Mr Robert Hunt), Roles & Economic Considerations (Mr Reginald Hillsdon), and Design Synthesis (Mr Brian Wilson). Gabriel A. Khoury J. David Gillett
Xli
Acknowledgements The editors wish to express their appreciation to the other authors of the different chapters for their co-operation and valuable contributions. The contribution of The Airship Association - direct and indirect - is also very much appreciated. Mr Dick Chadburn, Chairman of the Association, has kindly granted permission for the use of diagrams from the Association’s publications. The assistance of Westinghouse Surveillance Ltd. in furnishing information and diagrams on the Skyships series of airships for a number of chapters in this book is gratefully acknowledged. Contributors of other graphical material are also gratefully acknowledged.
These
include The
Imperial
War
Museum,
Mr Nigel Wells,
Dr J.
Bracher, Dr Masahiko Onda, Zeppelin Luftschiffstechnik, Dr Edwin Mowforth, and Professor J. DeLaurier. A special note of thanks is reserved for the professional skills of Mrs Valerie Till who has re-drawn nearly 200 line drawings and graphs. This has contributed so much to the consistent appearance of the book and to the high quality of its diagrams. Last but not least, we thank our wives Adriene and Irena for their patience during the many years of preparation of this book.
POSTSCRIPT Sadly, my co-editor and good friend Professor J. David Gillett OBE, DSc died on the 14" of October 1995 aged 82 before publication of this book. Professor Gillett was a distinguished scientist, a former Pro-Vice-Chancellor of Brunel University and President of the Royal Entomological Society. From his early days, Gillett was intrigued by airships, having observed the giants of the sky including the British R33, R34, R36, R100 and R101 and the German Graf Zeppelin. Gillett was a founding member of the Airship Association (1971) and served with me on its Technical Committee. He was elected member of council and Vice-Chairman of the Association, in which capacities he served until he died.
xill
This book is dedicated to Mr Roger Munk for his contributions to the design and development of the modern airship
char
fay!
AOA ‘US
oer Tee
es
«
fo
us
0
~
ria Vk
ae
~————s ~>
;
TRL «| (2
4
ia
A
Yv ecioe | -tecnuneneigs
tie Mada ile, eyches ives ahead
~
ane, Prue? ate
A "irae pi SF
4°
tot Val a
_
ig 2 creda
&
Se
7
a. iePe
_ 4
es
we
a.
i
oo
-
;
Rear
7
ri
#7 >
—
jreapel iat Dye “a
na?
—
6 Pee
Sen
atni
Tn
jest
fe
mS
a?
S oa
he
ee
es
7
ates
habe
rn
a
ib
pe
4
ri
of
2
Qa
|
:Oe
Gleenee
‘Soe
to
poe
us
iG ‘aaa
a Akl
@hipees ae
grams)
Ay:
ee
a
cag(ries is
eta
ve
be
‘
De Raeeipe
@
a2
=e re
O40:
PY cee
a
e. (eee rem!
i
7
inner
epmes
oe
ae
a
. a
ie
°
a
:7
tg gett be si Th
*
7
_
@@®8|
Ne 2 Rede PP fig a te ewe pallies ega fe
dit
earn
es
ee
1,4
‘Map!
Mine Opes
ssa
Gia?
ow
be
xing so NE rare Sealsae oem V then ps< Pa, The force on the small area around this point, relative to that which acts when the body is at rest and the local normally static pressure is pa, 1S (ps-paJOA, which is negative and therefore acting
outwards. Non-dimensional pressure coefficients allow comparison between different flow conditions and/or shaped bodies. The non-dimensional pressure coefficient is defined as:
Cp =
(ps-pa)/0.5 p v2
(3.2)
26 Airship Technology
C, has the value 1.0 when V, = 0, i.e. the flow is brought to rest on the surface. It Hecemne: negative when V, > V. An example of the pressure distribution around a circular cylinder is shown in Figure 3.1 curve A. From the symmetry of the pressure distribution over the front, from 270° to +90° and the rear, 90° and 270°, it is clear
that there will be no net force in the direction of the free stream, i.e. no drag force. The reason for this ideal, and not realisable, result is that viscous forces have been
neglected. Air has a very low viscosity; at 300 °K it is 18.325 x 10-¢ but that for water is 1.0 x
10-3 Nsm~2. Nevertheless the value for air is sufficient to produce a shearing action in
the flow along the surface of a body. The air in immediate contact with the body is at rest relative to the body. In laminar flow consecutive layers of air move parallel to the surface at increasing speeds until the inviscid solution speed is attained. However, this orderly arrangement eventually breaks down and the flow becomes turbulent. Eddies are produced which entrain fluid from the free stream and mix it with the sluggish air already there. This region of lower velocity air in both laminar and turbulent flow is called the boundary layer. The loss of air velocity in the boundary layer results in a loss of momentum
and hence a drag force, which is naturally called skin friction
drag.
eed
(OO
ae
[Tf
BAVA AN? a
se
/ Incident flow
O
90
180 Angular
Figure 3.1.
position
270 © (degrees)
Pressure distribution around a circular cylinder.
360
Aerodynamics
27
The development of a boundary layer is affected by the pressure on the body. Figure 3.1 shows that the static pressure decreases over the nose region 0° to 90°. This pressure gradient accelerates all the fluid including the boundary layer. The thickness
of the boundary layer therefore grows more slowly. However, in the region 90° to 180° the opposite effect is observed resulting in a deceleration of the fluid. The fluid close to the surface already has a low velocity and this deceleration causes it to move in the opposite direction to the local free stream. When the net local velocity in the inner boundary layer becomes zero, conservation laws mean that the fluid moves
outwards from the surface. The boundary layer therefore separates from the surface and a wake is formed (Figure 3.1, continuous curve). The static pressure on the surface of the body within the wake is no longer governed by the ideal local velocity but remains roughly constant at the value where the boundary layer separated. The symmetry of the pressure distribution over the front and rear of the body is destroyed and a drag force results. This contribution is called form (or pressure) drag. Form drag can be altered by streamlining the shape of the rear of the body. The shape is chosen to reduce the rate at which the local static pressure rises in this region. Aerodynamic lift is the component of pressure normal to the free stream. Total lift is obtained by integrating this component over the whole body. For both curves in Figure 3.1, the symmetry of the upper and lower surface pressure distributions results in zero net lift. To produce lift the local flow velocities must be greater over the upper than over the lower surface of the body so giving rise to different static pressure distributions. Such a difference in pressures is usually produced by flying the body at an angle to the free stream. This change in pressure distribution affects the boundary layer development and hence the drag components discussed above. The preceding discussion assumed that the pressure distributions measured in the direction of the free stream were identical for all stations at right angles to it (chordwise and spanwise for an aircraft wing). All real lifting bodies are finite. At the ‘spanwise’ limits of a lifting body the pressure difference between upper and lower surfaces cannot be supported and air flows from the lower to the upper surface. This results in a trailing vortex pattern; an example being the vapour traiis of high flying aircraft. The trailing vorticity produces a velocity on the body normal to the free stream which tilts the incident flow slightly downwards. The lift vector, which is normal to the incident flow, tilts backward by the same amount. Resolving the lift
force along and normal to the true free stream produces a component force in the drag direction. This is called the induced drag because it is induced by the production of lift on a finite size body. Typical airship hulls experience all of these effects. It will be
readily appreciated that the three-dimensional flow around such a body, when it is inclined at an angle to the incident flow, can result in a complex shed vortex wake and associated cross flows on the body.
The above discussion has concentrated on forces on a body but the local pressure forces produce moments about reference axes which play a major role in determining
the control requirements of a flying vehicle. These will be discussed further as the need arises.
28 Airship Technology DRAG
A given lift is produced by an aircraft wing which has a much smaller volume than an airship. However, the wing has to be in forward motion to develop lift. Drag is therefore of primary importance. The Bare Hull
It is advantageous to express all aerodynamic forces and moments on a body in non-dimensional form. Aerodynamic forces are divided by the dynamic head of the incident flow V and a reference area A. Thus the lift C, and the drag Cp coefficients are expressed as:
C. = Lifv/(0.5pV2A)
Cp = Drag/(0.5 p V2 A)
In these two expressions only A is not unambiguously defined. The lift of an aircraft is produced almost entirely by the wing so the relevant area A is the wing plan area. The lift of a buoyant craft is related directly to its volume so A is often chosen to be
(Buoyant Volume)?/3 However, owing to the large surface area of buoyant craft, skin friction is the largest fraction of the total drag. For this reason many reports use force coefficients in which A is the surface area. Care must always be taken when comparing non-dimensional coefficients taken from different publications to check that the same variables have been used. It is possible to determine a conversion factor between the two non-dimensional forms for a particular class of shapes. For bodies of revolution of length 1, maximum diameter d and with a fixed volume:
d?.1 =
constant
Young (1939) calculated for bodies based on the R101 shape the variation of A/I? as a function of d/l (The ratio I/d is the Fineness Ratio of a body. Using the ratio d/l, however, produces simpler approximate formulae for drag and the like. The ratio d/l is the Thickness Ratio. Although d/l is a ratio it will be expressed by the value of d as a decimal of 1). The result, for 0 < d/] Q-6 laminar flow x =0:046 x=0:258
Reynolds number =10 Body thickness ratio qi ; 0
0-2
0:4
x
Fraction
Figure 3.2a.
0-6
0:8
1
of body length
Local skin friction distribution.
0-0018
:
Cr = Skin friction
0-0006
Non-dimensional
Figure 3.2b.
turbulent
transition
point behind
Variation of total & skin friction drag with transition point position.
nose
32 Airship Technology
Curve
Transition
A
Point
x/l = 0-0 (Nose) x/l=0-2 x/l\= 0-4
/Total drag Form O
0:06
Figure 3.3.
0:12
d/I
0-18
Thickness
ratio
0:24
0:3
drag/Total Skin friction drag
Variation of form & skin friction drag with thickness ratio.
Hoerner (1957) analysed experimental data and compared them with theory to produce empirical formulae which may be used to assess the drag of bodies. In his section 6c he deals with the drag of streamline bodies. For most airships the flow over the hull is turbulent and for these conditions it was deduced that:
CoC, = 4(I/d)'¥3 + 6(d/)2 + 24(d/1)2.7 Equation 3.5 has a minimum value when the thickness ratio d/l = 0.217.
(3.5) Loe)
This is
greater than Young's figure of 0.182 but both curves are very flat in the region of the minimum. Hoerner suggests that C; for practical levels of surface roughness and for Re > 5
x 10° varies as: C; =
0.043/Rel/6
(3.6)
Combining Equations 3.5 and 3.6 gives:
Cov =
{0.172(1/d)'3
+ 0.252(d/1)!\2 + 1.032(d/1)2-7}/Rel/6
(3.7)
Cpy has been calculated using Equation 3.7 for the values of d/l in Table 3.1 and the
results, together with those by Young, shown in Table 3.2. Young's work is largely
theoretical while Hoerner uses mainly experimental evidence; the agreement is thus acceptable.
Aerodynamics
Table 3.2.
33
Values of Cpy given by Young and by Hoerner for typical airship hulls
0.05
es Young
The National Physical Laboratory in England suggested a low drag shape for an airship hull. A cross section along the axis of symmetry was defined mathematically. The shape with the defining equations is shown in Figure 3.4. The radius of curvature increases from bow to stern, thus reducing the chance of flow separation. Reid (1987) has pointed out that the later Skyship designs bear some similarity to this shape, although the point at the stern is rounded off as inflated structures tend to droop if pointed. The bare hull only represents part of the airship drag even when flying with no lift or yaw. The effect of other components on total drag is now considered.
2
2
2
Le a?
2
os Gy De Ee b?
2at
b*
S eI
Figure 3.4.
Ae el ode
anaes
The NPL low drag airship body shape.
Effect of Additions to the Bare Hull
To make an operational airship it is necessary to provide control surfaces,
of these items propulsion unit(s) and accommodation for the crew and payload. Each it is fitted to when has a drag when measured on its own. This drag will be modified Conversely, each the hull as it is then operating in the flow field of the hull. its aerodynamic forces. component will affect the flow around the hull so modifying
made for each design. Wind tunnel tests and fluid dynamic calculations have to be
34 Airship Technology
Durand (1934) contains component drag data for three airships which are reproduced as Table 3.3. The drag of each group of components is expressed as the flat plate area with the same drag. It also shows each component group's drag as a percentage of the total drag. The bare hull contributes about 50% of the total drag. While low hull drag remains an important objective, the minimisation of the drag of the whole design is the final goal. The data in Table 3.3 were carefully estimated in a variety of ways but it is worth repeating Havill's observation verbatim. "By attributing different degrees of importance to the various component data derived from indirect evidence, conclusions can be shifted somewhat."
Table 3.3.
Drag Breakdown for three airships
Estimated Drag Area Breakdown
Los Angeles
:
Without Water Recovery
Macon
a Bare hull Fins and rudders
aa~—
:
Wing power cars or outrigger gears, their
0
:
33.2 | 39.6
12.5
:
14.0
— oN 0
6.8
E.Gup
suspension, ladders, struts, hoods, radiators, exhaust mufflers.
Rear power car with handling rails and bumpers. Control car or passenger car with handling rails and bumpers.
Miscellaneous protrusions - Mooring mast equipment, hoods etc.
Or a + cof is
Nn Ny |SNNi]
20. =) (Volume)? Resistance Coefficient C
~~\oSo
0.025
100 | 41.
Oo
oooaN
0.022
LOY
fF W527
Aerodynamics
To illustrate the modification produced from data contained along the top and bottom of a without gondola. Curves C & virtually identical. Curve B is
35
to the hull pressure caused by the gondola, Figure 3.5, in Gomes (1989c), shows the pressure distribution wind tunnel model of the Sentinel 5000 hull with and D are for the clean hull; at zero incidence they are for the top of the hull fitted with the gondola and. as
expected, there is little change from curves
C & D. However, curve A for the lower
surface where the gondola is fitted shows considerable change. As might be expected the pressure rises ahead of the gondola. Unfortunately the two static tappings nearest
to the gondola were blocked so the shape of the pressure curve over it is not known. Aft of the gondola the flow accelerates back to a slightly higher speed than the without gondola case. The increase of static pressure over the gondola will contribute to the aerodynamic lift of the airship. There is no evidence from this case that the gondola has affected the flow at the rear of the hull.
1
ee a
ea
ee
A
—
Goncolas
On@
C
0.
B “\ :
een Soe N.B.
eee
= D
Location of rows of pressure tappings_ Pressure taps at 0:4 & 0:6 blocked on
curve
A
0-1 9 oe,
-0:5
—
= length
0 Nose
at 0-4
0-2
0-6
Non-dimensional
at
0-8
1
hull length
Figure 3.5. _Effect of gondola on pressure distribution of Sentinel 5000 airship. Curtiss. Hazen and Putman (1976) have compiled information on tai! drag compared with bare hull drag; this is copied as Table 3.4. The bottom line of the Table shows the considerable variation in the results. This may be due to the accuracy of the methods used to obtain the data, but it also must reflect the differing aerodynamic
interference which occurs. As a first approximation the regression line to the data in the bottom line of Table 3.4 is:
Tail Drag per unit area Hull Drag per unit area
=
3.56
- 0.195(I/d)
(3.8)
36 Airship Technology
Table 3.4.
Tail drag compared to bare hull drag
Thickness Ratio
0.149 | 0.098
Wetted area of Tail Wetted area of Hull
Drag of Tail
0.122
0.08
COTS
0.27 | OA8..|
0.07 | O12 | O07.)
OOTs
Ont
O22 ) O24
Drag of Hull
drag/Unit Tail area Hull drag/Unit area Bod. = Bodensee,
L.A. =Los Angeles,
§ Akr. = Akron
Attaching two bodies together in an airflow results in the flow about each being modified. A fin on an airship results in the local velocity over the hull, outside the boundary layer, being increased due to the flow blockage of the fin. This leads to an increase of skin friction drag and may also affect the local pressure distribution. Figures 3.6 & 3.7, also from Gomes (1989c), show the pressures on the hull in the
vicinity of a fin.
Fin at 45°: pressure tappings at 22.5 and 67.5° No fin: pressure tappings at 22.5, 45 and 67.5° Fin and pressure tappings at 45° and x = 0.7 and 0.95 only.
0:7
0:76
0-82
x
0-88
0:94
1
Non-Dimensional Hull Length
Figure 3.6.
Effect of undeflected ruddervator on hull pressure distribution.
Aerodynamics
37
The fin is aligned along the line through the 45° point on the central meridian. The three curves in Figure 3.6 labelled B relate to the bare hull. As expected for an axisymmetric body at zero incidence they are virtually identical. Adding the fin blocked the static tappings between x = 0.8 and 0.9 so curve A only shows that the pressure up and downstream of the fin is similar to that of the static rows shown in curve C. The reduction in static pressure, indicating a higher velocity, due to the fin is apparent by comparing curves A & B. This velocity change will increase the form drag as the slope of the hull gives the force a rearward component. The change of velocity gradient may also affect the drag by altering the boundary layer separation. Figure 3.7 shows the further change in surface pressures when the control surface is
deflected. The increase in static pressure for curves o & a and the reduction for curves c & d in the right hand relative to the left hand diagram is due to the lift induced circulation around the ruddervator. This term indicates that the control surfaces on the fins act as elevators when all are deflected in the same sense and as rudders when the deflections of diagonally opposite pairs are in the opposite sense. Owing to the length of airship hulls, the boundary layer is thick at the stern. Hoerner (1957) provides formulae for the thickness of the boundary layer s (i.e. distance from the surface at which the velocity is 99% of local inviscid free stream value) on a body of revolution aligned with the wind. If the maximum body diameter d occurs a
distance x,, behind the nose, then: s S
=
0.02x
for
x = Xia
0.02x(d/d.)«
for
x >X,
(3.9)
where: d,, is the diameter of the body distance x behind the nose. The constant of 0.02 was derived for aircraft fuselage surface finish quality: changes from that standard will alter this value. Within the boundary layer the reduced velocity reduces the drag of the fin. For the YEZ-2A this formula gives a value of s in the mid-fin position of about 3m which is nearly half the height of the fin. However, the displacement thickness (i.e. the thickness of a layer of stagnant fluid on the surface that has the same integrated velocity defect as the actual boundary layer) 1s about 1/4 of s or in the case of the YEZ-2A about 12% of the height of the fin. Estimating control car drag may be done using information in Hoerner (1957) which is based on experimental tests of cockpit shapes. His results suggest a drag coefficient of 0.09 based on the frontal area of the car. He does qualify this with the statement, "Taking into account, however, the gaps, edges and other surface irregularities of the canopy as actually constructed, its drag is probably twice as high as found on smooth wind-tunnel models." The value of the velocity used to convert drag coefficient to
drag force is the local velocity around the hull.
No published breakdown of drag exists for a modern airship. However, the total airship drag coefficient of a Skyship 500 was experimentally determined from flight
given in tests as 0.032 by Bailey (1985). This value is significantly higher than those Reynolds lower a at operates Table 3.2 but the Skyship is smaller and therefore number.
38 Airship Technology
Rudder setting 0 degrees
Q-2
No pressure tappings between 0:7 & 0:95 due to attachments
on curve
2
Oi?
OG
b
| 0:82)
Non-dimensional
Rudder setting
086)
0-04
1
hull length
25 degrees down
e) Viewed from rear
07-2
-0°76)-.0'62" 1O'882410,04 Non-dimensional
Figure 3.7.
1
hull length
Effect of ruddervator deflection on hull pressure distribution.
Aerodynamics
39
The bare hull drag coefficient calculated using Equation 3.7 is Cpy = 0.017, ice. just under half of Bailey's measured figure. This is in good agreement with the examples quoted in Table 3.2. The drag of the tail calculated using Equation 3.8 increases the hull figure by 29%. The drag of the control cabin using a Cp = 0.18 increases Coy by 0.0064, which gives a total Cov = 0.0283. There are further contributions to be added including propulsion units, landing gear and mooring attachments. The agreement with Bailey's figure is acceptable.
DYNAMIC FORCES A body in motion, even in an ideal fluid, displaces that fluid. Each fluid particle
is displaced in the direction of the motion of the body. The magnitude of that movement is related to the distance from the path of the body. The fluid therefore gains kinetic energy which can be calculated using potential flow theory. The reader who is unfamiliar Thomson (1955).
with this theory is referred to standard texts such as Milne-
Apparent Mass Determination
If w is the complex potential of a flow then the kinetic energy in the fluid Ty is given by:
(3.10)
T,=-0.25 ip [wd W where:
iis
V—1 and w is the complex conjugate of w.
For the case of a circular cylinder of radius a:
w
Ualz,
=
Ww =
Uasz,
dw
=
Uadz/z?.
On the cylinder z= a e/9, Substituting these expressions into Equation 3.10 gives:
(3.11)
Uni pant
Tee.
The mass of fluid per unit thickness displaced by the cylinder is M' = 7 p a’. If M is the mass of the cylinder per unit thickness the total kinetic energy of fluid and cylinder T is:
05(M+ MU?
T =
(G2)
Let F be the resultant external force on the cylinder, so: FU
=
dT/dt
=
(M+M'U dU/dt
(3213)
40 Airship Technology since the rate of change of kinetic energy equals the work done by the external forces. Rearranging gives: F - M'dU/dt
=
MdU/dt
(3.14)
In the absence of the fluid the second term would be absent and the normal dynamic equation of motion results. The cylinder experiences a resistance to its motion due to the fluid of M'dU/dt per unit thickness. M' is called the added mass. It will be observed that if U is constant then there is no added mass. The physics behind this process was best explained by Darwin who is quoted in Milne-Thomson (1955). The following is based on the relevant section of that book. Consider the steady motion of a circular cylinder of radius ‘a’ moving through a perfect fluid of density p with velocity U. Figure 3.8 shows the motion of three fluid particles relative to a fixed observer as the cylinder moves from -o to +00.
Figure 3.8. _Fluid particle paths during passage of circular cylinder.
Aerodynamics
41
The equation of the streamlines is:
VR
=
4(y-v/2)/a2
(3:15)
where: R is the radius of curvature and v the streamline variable. When the cylinder is at -00 all the particles lie on the broken curved line on the left of the Figure. When the cylinder reaches +co they lie on the right hand broken line. The numbers on the curves show the particles’ position when the cylinder has advanced that number of radii from the central position. It is obvious that a mass of fluid has drifted towards the right of the diagram during the cylinder's transit. This ‘drift’ for any particle is evaluated by considering motion relative to the cylinder regarded as fixed. The drift € is then:
& =
(3.16)
|(U+dx/dtdt
This integral can be evaluated in terms of elliptic functions. For unit thickness the mass of the fluid which has drifted is:
p|_S4v where: v is the streamline variable. On evaluation the drift mass equals the mass of fluid displaced by the cylinder which is the same result found in Equation 3.11 above. The importance of using the apparent mass (M + M’) in place of physical mass M depends on the relative densities of the cylinder and the fluid. For buoyant craft these densities are very similar and apparent mass is important. The use of the complex potential to obtain the added mass of a body has been used for two- and three-dimensional shapes. Milne-Thomson (1955) gives results for the following bodies. For an elliptic cylinder with minor axis 2b and major axis 2a which is inclined to the
flow at angle a, M' = px(b2cos?a.+a?sin’q). It will be noted that this corresponds to
the area of a circle with a diameter equal to the projected width of the ellipse normal to the incident flow.
For an ellipsoid with equation x/a? + y*/b? + 27/c? =
1,
M' = 1.33pmabe op/(2 - a9)
(3.17)
where:
C=
(b?+A) 0.5 (c?+A) ost fp abc da / {(a2+a)!’
for an ellipsoid The general solution is evaluated using elliptic integrals. However,
, then: of revolution when b = c, which is relevant for airships
42 Airship Technology
Oy = 2{(1-f2)/f} {0.5 Ln[(1+f)/(1-f)] -f} where: f = (1 - b2/a2)®-5. In the limiting case of the sphere, when b = a, 1 = 2/3 so the added mass is half of the mass of the fluid displaced. If the ellipsoid is moving parallel to either the y or z directions then expressions similar to Gg exist. An expression may also be found if the ellipsoid is rotated about its z axis, i.e. it is yawing. Munk (1924) applied this information to determine forces and moments on airships. {Readers may find it easier to locate his work in his contributions to Durand (1934)}. His method was to modify the body mass with an inertia coefficient to account for the added mass. He examined both two- and three-dimensional problems. For two-dimensional cases, such as sections of the hull in a cross flow, the shape is approximately elliptic. Munk proposed that n be defined as:
n =
nc2/4S
(3.18)
where: c is the largest width at right angles to the motion considered and S the area of the cross section. If the section is circular n = 1. He points out that even in the case of pear shaped bodies the error in this approximation is small. The area of apparent mass is nS. The apparent mass of the hull is affected by its geometry with the shape of the bow and stern having most significance for axial motion. This is not surprising as the disturbance of the surrounding fluid reduces as the shape becomes more streamlined. Hence the drift mass reduces. Munk suggested two methods of force estimation for three-dimensional bodies.
The first approach resolves the incident flow along the body principal axes. For lateral motion the body is then divided into slices parallel to the incident flow component. The inertia factor for each slice is then calculated using Equation 3.18. The overall factor is then obtained by integration. The second method fits the hull shape with an ellipsoid of revolution with the same meridian cross section area S' and length L as the hull. The thickness ratio is then:
b/a =
4S'/nL?
Inertia factors, k, ,k2, k’ of this ellipsoid in axial, lateral and rotational motion about
the z-axis, yawing motion, may be calculated generally. For axial motion the relevant formulae are given at Equation 3.17. The equivalent expressions for the lateral and rotational cases may be deduced similarly or found in Milne-Thomson (1955) sections 16.52 and 16.53, page 485. The values for the inertia factors for ellipsoids of revolution are given in Table 3.5. If the hull is not circular in cross section then the appropriate inertia factors are nk,, nk, and nk’ where n is given for a cross section by
Equation 3.18 or by its average value over the length.
;
Aerodynamics
Table 3.5.
43
Inertia coefficients of an ellipsoid of revolution
Thickness
Direction of Motion
Ratio
Transverse
The Determination of Body Forces and Moments
If a body is moving steadily through a fluid, its motion may be defined by translation U and rotation @ vectors related to axes fixed in the body. If € is the momentum and A the moment of momentum vectors, then the normal equations of motion are:
se/et = oXE +X SVEt = wOXA -UXE
+L
(3.19)
where: X and L denote an external force and moment respectively. Flight at Incidence
Consider a body of rotation flying steadily at incidence a. If unit orthogonal side and vectors i, aligned along the axis of symmetry of the body with jand _« to the upward respectively then: Wa:
ui
iwi
Ulicosa) 5 _Ksin @)
If the apparent mass in the two vector directions is A and B then:
44 Airship Technology
—& = AUicosa + BU_xksina Substituting into Equation 3.19, noting that the left hand sides are zero as the motion is steady and that @ = 0, then:
X
=
OL
=
UXE
= (B-A)U*cosa sina j
(3.20)
No force is generated on the hull but there is a nose up pitching moment, if @ is positive, which will rotate the hull until the axis of symmetry is vertical. But A = p(Vol)k, and B = p(Vol)x2 so the magnitude of the nose up pitching moment M is:
M = %p U2(Vol)(ky-k,)sin 2c
(3.21)
where: (Vol) is the hull volume. This pitching moment has to be countered in steady flight and this is done by the tail surfaces. These produce a lift force equal to M/I, where |, is the distance from the centre of lift of the fins to the centre of the hull. The airship, therefore, always has some dynamic lift if the hull is flown at an incidence, although the inviscid flow model shows - as expected - that no lift is generated on the hull. Steady Turning Flight
An airship flying a steady circular course of radius r is shown in Figure 3.9. The centre of gravity of the hull moves with velocity V along the tangent to the flight path; the hull centre line flies with an angle of yaw, 6. To maintain this attitude the hull has
an angular yaw velocity of Vcos ¢/r. Thus the velocity components U, and U, of cross section area § at position x are:
Ux =
Veosp,
and
Uy =
‘
Vsinod + xVcos$/r
The rate of change of momentum per unit length of the body is:
d/dt(pS.Uy) = Ux.d/dx(pS.Uy)
=
pV?sin 2pdS/dx+ (S+xdS/dx)pV2/r_
(3.22)
The force distribution along the hull corresponding to each of these terms is shown in
Figure 3.9. Curve b is the distribution that gives rise to the moment discussed in the section, Flight at Incidence The moment here tends to tighten the radius of the turn
and rudder(s) deflection is needed to maintain the constant radius. In this case the ‘incidence’ is the yaw angle. The second term, curve c, represents the aerodynamic forces combating the centrifugal acceleration owing to the curved path flown while curve d is the equivalent of the first term but the ‘incidence’ is caused by the angular velocity of the body. It is worth noting that the incidence of the rudder(s) is increased
by the angular velocity of the hull which enhances their effectiveness.
Aerodynamics
Constant section
1
®
x
=
Stern =
Radius—. a. Attitude
b. Pressure
STC c. Pressure
d. Pressure
Figure 3.9.
of airship to path
distribution
due
to yaw
LU CUE se
distribution to combat
distribution
=
centrifugal force
due to angular velocity about
mid-point
Forces developed on an airship in curved flight.
45
46 Airship Technology
The inherent instability of the airship is never completely corrected by the fins. This is partly for reasons of fin weight and size, the latter for ground handling and hanger size, and because an airship turn may be induced more easily as the instability swings the nose into the turn. The penalty is that the helmsman, or flight control system, has to correct any disturbance induced change of heading. It has been found that the aerodynamic forces increase more rapidly than linearly with @. Thus with rudder zero the airship settles into a small but steady rate of turn in the direction induced by a disturbance. There is therefore a hysteresis effect in the response to rudder movement about zero deflection which is shown in Figure 3.10 taken from Arnstein and Klemperer (1934).
1/R
Curve a...a 5.220
Airship LZ126 R32
a ~15
-9
—3
3
St’bd Rudder
angle
9 Port
15
(degrees)
Figure 3.10. Airship path curvature vs. rudder angle.
SLENDER BODY THEORY
In calculating forces on a body of revolution, like an airship hull, Munk made the assumption that it was slender. A body is slender if the length is large compared
with the maximum diameter and the rate of change of diameter along the length is small. The last condition allows the axial component of velocity at all sections to be
assumed to be identical to the component of the free stream. The cross flow is then
Aerodynamics
47
two-dimensional in the plane normal to the long axis. For a circular cross section the pressure distribution can be calculated easily in inviscid flow. For any circular section the velocity potential © is:
Oo
=
“Uy + R2/r)cos 0
where: Uy is the cross flow velocity, R the radius of the cross section and r,6 the polar
co-ordinates with the pole at the centre of the circle. If B is the slope of the surface in the axial, x, direction, ie. tan B = dR/dx, then the pressure p at any point on the circular section is: (p-Po)/p IfP
=
=
le 2UxUy(R/r)tan b cos q sin 2a-0.SUy?(1-4sin? q)sin° a
(3.23)
(p- po)/p, then Equation 3.24 may then be written as:
P =
Pg= ot 2tanB sin 2a cos 0 + (1-4sin2 @)sin2 a
(3.24)
where: P, = 9 is the pressure distribution on the circular section at a = 0. Owing to body incidence the force, f, on the cross section per unit length is given by: f =
(P-Py=o)Rcos 0d8 =
qo2nR(dR/dx)sin 2a
=
qodS/dx sin2a
(3.25)
where: S is the area of cross section of radius R. Ward (1949) showed that this force is directed midway between the normal to the axis of revolution and the normal to the wind direction, i.e. a/2. Equation 3.25 is therefore multiplied by cos @/2. Allen and Perkins (1951) extended this analysis to make allowance for viscous effects. Using Equation 3.24 they computed the pressure distribution around several sections along the hull which were then compared with experimenta! data measured by Freeman (1932) on a 1/40th scale wind tunnel model of US airship ‘Akron’. Figure 3.11 is copied from their work. Flow separation occurs in an adverse pressure gradient and that is clearly shown in this Figure. Over the forward sections of the hull, x/L = 0.111, where the pressure gradient is negative, calculated and measured results agree well. Aft of the centre of the body an adverse gradient exists and flow separation may occur, any separation becoming more
severe the nearer the section is to the tail as
shown at x/L = 0.867. Allen and Perkins realised that if the body pressure distributions were viewed sequentially from nose to tail, there was a strong similarity to the time history of a circular cylinder's pressure distribution as the flow is accelerated from rest. In that flow regime vortices were shed behind the cylinder; photographs of such a flow are shown on plate | of Milne-Thomson (1955).). Viewed three-dimensionally the vortex elements from each body section combine to form line
vortices along the length of the body. Allen and Perkins found flow visualisation photographs taken on missile and other slender bodies showed these trailing vortices. Since details of this flow and the drag coefficient are Reynolds number dependent
48 Airship Technology they suggested an additional term be added to Equation 3.25 to approximate for these viscous effects. This term is the drag of a circular cylinder at the Reynolds number Re for a stream velocity of Usin a and a diameter of 2R at lengthwise location x. Equation 3.25 becomes:
f =
qodS/dx sin 2a cos a/2 + 2R Cp (Re) do Sin’ &
where: Cpe) is the appropriate drag coefficient.
P-P Coefficient Pressure Incremental
P-P Coefficient Incremental Pressure
Body cross-section angula position 6° Figure 3.11.
Comparison of theory and experimental incremental pressure coefficients for U.S.S. Akron.
(3.26)
Aerodynamics
AN ESTIMATION METHOD FORCES AND MOMENTS
49
FOR OVERALL AERODYNAMIC
Equation 3.26 has been used by Jones and DeLaurier (1981) to provide a method for estimating airship forces and moments. They followed Munk in noting that airship hulls have conicity, which means that the surface pressures on a crossflow section result in both lateral and axial forces. Equation 3.21 allowed for this effect with the factor (k, - k,) so the first term in Equation 3.26 becomes: QodS/dx(k; - k,)sin 2acos a/2.
Their schematic of the airship steady state analytical model is shown in Figure 3.12. The hull is considered from the nose to the start of the fins, l,. Aft of that station it is included with the fin aerodynamics. Hull forces are evaluated on a typical section
situated between 1; and 1;,, using Allen and Perkins and then summed. Fin forces are estimated using one of a number of well established finite wing methods. Jones and DeLaurier (1981) use Wardlaw (1979) for values of (Cyc), and (Cog)g. (Cpe is a correction to Wardlaw as the fins do not have sharp leading edges. These results are multiplied by efficiency factors, n, to allow for the effect of the fins on the hull and n¢ the effect of the hull on the fins. These factors have been obtained empirically by a least squares fit to experimental data.
Figure 3.12.
Schematic of steady-state analytical model.
50 Airship Technology
The normal, C,, and axial, Cp, force and pitching moment, (Cy)nose» Coefficient
equations are:
Cy =
[(-*) 1k 1,+0.5(Cna” ene S¢ sin 2a +
[(Coodh J+ Coe)e Si] sin ot sin fo
(3.27)
Co = [(ConoSn + (Cop)o Se] cos?aN, [sin 2a sin (a/2)-(Cy)pS (ky-k,
(Cr)nose
(3.28)
ene Splsin 20 — = > [o-Ay) nk 4t0.5 (lp), (Cra
(Codpys + (Code (py S41 sin @sin jo!
(3.29)
where:
e
Cy Cp are made non-dimensional with qo(Vol)*?
e
(Con)nose with qo(Vol)23L, where L is the total hull length
e
Se
e
SS, =
e
S
o
Tt, =
=
fin(s) reference area S; (fig 13)/(Vol)*? hull reference area /(Vol)*?
= _ hull cross section area at station €
oY,
fF dSMEdEL Sree
= Jy =
ff EdSMdE dé (VolPSL fo 2r& d&/(Vol2BL
e = (7p), and (/,))_ with L e
§(Coo)h
=
hull cross-flow drag coefficient //,L,
oe . (Cade
=
fin(s) cross-flow drag coefficient /S¢
©
(Cy)p
=
derivivative of isolated fin's normal force coefficient w.r.t. o @ a=0/S¢
©
(Cpr)o
=
hull zero-angle axial drag coefficient /S;,
©
(Cpp)o
=
fin zero-angle axial drag coefficient /S¢
e
(Cyr
=
fin leading edge suction coefficient /Sr
The integrals depend only on the design geometry.
Aerodynamics
5\
Sth
Figure 3.13.
Definition of fin planform areas.
The fin geometry is shown in Figure 3.13. The carry-over of lift from fins across the dividing hull has been appreciated since 1910 when Prandtl and Fuhrmann showed that two fins separated by a body produced 60% more lift than the two fins directly joined together. Figures 3.6 and 3.7 discussed above show the effect. Data obtained on the model of the ‘Akron’ and separately on a model of the R101 suggest a slightly lower figure as illustrated in Figure 3.14, taken from Curtiss, Hazen
and Putman
(1976). Nk and nf are shown in Figure 3.15, taken from Jones and DeLaurier (1981). I is the dihedral angle of the fins. The interaction of the body vortices and those generated by the lift-producing fins can generate sharp local aerodynamic fin loadings. An example taken from the ‘Akron’ model tests of Freeman (1932) is shown in Figure 3.16. Jones and DeLaurier's model brings a logic into the estimation of aerodynamic forces and moments. It is capable of further development but this is hampered by the absence of a large data base against which the method may be tested and refined. Readers intending to use the technique should refer to the original paper. The steady aerodynamic loading distribution on the body may be calculated using the equations before integration. Such information is essential for the structural designer.
52 Airship Technology
Akron, Mk II fins
Re =19 x10° O Force,
(hull +fins —hull) © Pressure on fins only
Cir Tail Lift Coefficient
Angle of attack, a R101
Re=10° 0 Force,
(hull +fins —hull) o Pressure on fins only
Tail Lift Coefficient
Angle of attack, a
Figure 3.14. Influence of ‘carry-over’ lift on tail lift coefficient.
Aerodynamics
0
05 S_cos’(T)/()) total
53
1:0
Figure 3.15. Fin and hull efficiency factors deduced from experimental data.
54 Airship Technology
Upper
surface
Lower
surface
/
Hull
line
Se = 20° Elevator hinge
ary. Si en SUE
(ae
axis
Figure 3.16. Fin distribution on 1/40" scale model of the Akron.
2
7
Aerodynamics
55
UNSTEADY AERODYNAMICS The structural failure of airships, e.g. the Shenandoah, in conditions of severe atmospheric turbulence has led to the development of techniques to estimate structural loads in such conditions. These techniques are computational backed up with wind tunnel testing, the latter being discussed in the section on Wind Tunnel Testing. The computational method requires a model of the incident atmospheric turbulence and method of determining the vehicle's response. A characteristic of turbulent flow is that the pressure measured at a point is random in nature. Representation of such a flow is therefore more difficult than for steady flows discussed previously. The representation of the time varying pressure (or velocity) at a point is expressed as a series of sine waves each with a characteristic amplitude IT, frequency and direction in the appropriate frame of reference. Treating, for example, the vertical component of a gust Wo, the value at distance € from the origin fixed in the airship for gust component with frequency 1s: Wee
U, T exp{io t - ia § cos (4)/U,)}
(3.30)
where: py is the steady incidence and U, the airship velocity. The response to each component of turbulence may be evaluated, for example by inserting Wo/Ug, which is the local incidence at &, into Equations 3.27-3.29 above. The resulting forces and moments are then evaluated on the assumption that a quasi-steady model is appropriate. Owen (1973) gives an example of such an approach and also explains the results. Using slender body theory he computed the response of a mathematically simple airship-like rigid body. Since slender body theory only uses the variation of cross section area S, he specified the variation along its length] (0 w(X) [wi 0:4
Sm
@)
1071
1
10st
(b) Spectrum of the (vertical) w-
(a)
component of atmospheric turbulence
b HO/PTUZSE Ad IX)
> (x)/P2L2S2,, fw?
3:0
0-15
2:0
0:10
1:0
0-05
(a)
(c) 0
10-
7
1
10
—_—I
xf
10?
0
1071
1
10
xl
Note the sharp peak when the wavelength of the turbulent
velocity (2 77 /x) is equal to / the length of the hull
(c), (d) Spectra of the fluctuating lift on the hull.
With (w?)'? = 1 msec’ and U = 30 m sec’', the lift on the hull in root mean square is 0.06 % 9 U? Sp. The incidence required to produce an equivalent
lift in steady flight is more than 3°.
Figure 3.17. The response of an airship hull to atmospheric turbulence.
10?
Aerodynamics
57
If /= 330m then the expression fits data recorded during aircraft flights. The spectrum of the lift produced on the hull is given by:
D(x) = p?U,?By (x) S'(E).S'(n)-exp {ix (E-n)} dE.dy
(3.33)
where: S' = dS/dx. For the simple body shape in Equation 3.31. Owen evaluated
Equation 3.33 analytically. Figure 3.17¢ shows ®,(y)/p?U,2®,(y) as a function of xl. For the case when | =/ the spectra of the fluctuating lift are shown in Figure 3.17d. The airship length was chosen from a project design of Sir Barnes Wallace. If (w2)/2 = j ms"! then the root mean square lift on the hull is 0.06.2pU 9S im,which could also be produced by flying at an incidence of 3°. Figure 3.17c shows that the lift spectrum peaks when the wavelength of the turbulence equals the length of the hull. Owen explained this in Figure 3.18. The vertical velocity from the turbulence changes sign at the centre of the hull. The lift, however, is obtained by multiplying by S', which also changes sign at the hull mid-point. The lift on both halves of the hull is therefore positive. Other turbulent wavelengths will experience some lift cancellation. Any tail surfaces will produce a negative lift in the situation shown in Figure 3.18, which will pitch the airship nose up, so increasing the total lift. The actual gain in lift will depend on the period of the rapid oscillatory mode. DeLaurier and Hui (1981) use a similar approach but have taken the analysis further by estimating the life of an airship in various weather conditions. Using the method of caiculating aerodynamic forces and moments referred to in the section on Overall Aerodynamic Estimation, together with Von Karman's model for the turbulence
power spectrum (Press and Meadows,
1955) they used values of turbulence scale
length / and intensity o - obtained from flight test data combined with Etkin's (1980) mission analysis technique - and determined various designs' probable lifetimes. This analysis showed the large benefit to vehicle life if the speed is slowed upon entering strong turbulence. This agrees with the habits of Zeppelin captains on encountering strong turbulence, which no doubt contributed to their good flight safety record. DeLaurier and Hui have added a useful weapon to the airship designer's armoury.
58 Airship Technology
Vertical
(a)
velocity
Lift
(b) Generation of Lift by a spectral component of the turbulence with wavelength equal to the length of the hull. Note:
the usual gust ‘alleviation may be considerably weakened, even converted into a gust ‘augmentation’ through the action of the tail surfaces, which is to generate a pitching moment in such a sense as to increase the lift.
Lift on hull
eked tof ys
0:00
0:20
0-40
0-60
(A)
Wing-body
(B)
Wing- body Wing- body Wing- body
(C) (D) (E)
Empirical
0-80
results
1-00
(a) Fraction of fin covered by hull
0:96
0:00:
.0:10
0:20
0:30
(b) Effective fin area/hull
Figure 3.23.
0°40
0-50
projected area
Comparison of fin n; and hull 1, efficiency factors from CFD and experimental data.
63
64 Airship Technology
Figures 3.23a and 3.23b show the results of applying the full model to five hypothetical airships to calculate the efficiency factors n, and n¢ These are defined in the section on Overall Aerodynamic Estimation and are shown in Figure 3.15 as well as on this Figure. The five designs were produced by adding various fin arrangements to an ‘Akron’ (ZRS-4) class hull. The curves of Figure 3.15 were empirically deduced from data which included that for the ‘Akron’. The agreement is satisfactory and this provides hope that computational fluid dynamics will calculate conditions which have previously relied on empirical estimates. Wind Tunnel Testing
Wind tunnel tests are performed to answer particular equations. For performance the overall lift, drag and moments of the vehicle in non-turbulent flow are required. Forces and moments from control surfaces provide information to ensure satisfactory handling. Hinge moments from these control surfaces are needed by the control systems engineers. Detailed force distributions are used by the structural design group to optimise the structure. Separately from these tests the response of the vehicle to unsteady flow is required for reasons given in the section on Unsteady Aerodynamics. A brief summary of some of the methods and problems associated with such tests is given below. For a more complete discussion of the subject and test results for modern airships, the reader is referred to Gomes (1989a-d and 1990). Gomes (1990) is a summary of the work which is supplemented in the other references.
A basic problem with wind tunnel testing is that the model has to be attached to the tunnel structure. The attachment alters the flow around the model so care must be
taken not to affect the critical flow areas. Overall forces and moments of the body may be measured through the attachment, usually through a strain gauge balance recording up to six components. The only technique which does not require a physical attachment is magnetic suspension. The strengths of electro-magnets mounted in the wall of the tunnel are adjusted to hold the model in a fixed attitude and position in the tunnel. Model forces and moments are determined from the magnetic strengths. Most facilities which use this technique are small and, for reasons given in the next section, not suited to airship hull testing. Small force and moment balances may be included in components of the model, e.g. for control surfaces. General distribution of forces on a
model are obtained by pressure plotting the surface. Hypodermic tubing is inserted from inside the model to finish flush with the surface. The other end of the tubing is connected to a pressure sensor, which records the static pressure at the surface of the model. Owing to pressure lags in the tubing this technique is only practical for steady flow conditions.
Aerodynamics
65
Steady Flow Testing
To ensure that a wind tunnel test reproduces the full scale behaviour certain non-dimensional groups must have the same value for both model and full scale. These groups are determined by dimensional analysis. For steady flow simulation the variables usually considered are fluid velocity V, density p and viscosity 1, the speed of sound in the fluid a, the density o and modulus of elasticity E of the structure, a characteristic length | and gravity g. The non-dimensional groups which have to be maintained are V/a, p/o, E/pV2,V2/lg and pVI/u. The first group is the Mach, the fourth the Froude and the last the Reynolds number. The simplest wind tunnel uses air at atmospheric pressure and temperature. In this case p and tt have the same value at both model and full scale. Only in rigs like whirling arms can apparent gravity be changed so g is constant. It is then easy to show that the model has to be full scale to satisfy the equality of the groups. The choice then lies between the use of a more sophisticated wind tunnel, for example a pressurised or cryogenic facility, or to justify why certain groups are unimportant for the particular tests. For a slow speed vehicle like an airship, Mach number can be ignored as compressibility effects are absent. Since the model need not be buoyant neither p/o nor the Froude number need to be preserved. If E/pV2, the ratio of the elastic to the aerodynamic forces is large then a rigid wind tunnel model will be acceptable. This only leaves the Reynolds number to be preserved. In an atmospheric pressure tunnel preserving Reynolds number means that Valin =
Vr where m refers to model and f to full scale. In the Gomes (1989a-d and 1990) work the airship model was 1/75 full scale. So to preserve Reynolds number the tunnel velocity would have had to be 75 times full scale or about 2250 m/s. This would clearly make the neglect of Mach number a major error as well as severely limiting the number of suitable facilities. The Reynolds number is therefore not preserved but other techniques are available to minimise the error. Before discussing these techniques the limits to model size are explored. The size of the model must be related to the size of the tunnel working section so that the airflow around it approximates that experienced in free flight. If the model is large relative to the size of the tunnel working section then the air velocity around it will be higher than in free flight and the surface pressures lower. The aerodynamic forces and moments will then be in error. Consider the tests made by Gomes (1990) in the Cranfield College of Aeronautics 8x6 foot wind tunnel of 1:75 scale models of a
Skyship 600 and the YEZ-2A. For the latter model the maximum cross section is
about 3.3% of the tunnel working section area which means
an increase in static
pressure on that model section of 6.7%. Such an error is acceptable. For models with small thickness ratios, like airship hulls, which require testing at large pitch and/or yaw angles, the model size must be chosen so that neither the nose nor the tail approach close to the tunnel walls. For the YEZ-2A model, yaw and pitch settings up of to 30° were tested singly and together. For the 6 foot tunnel width the nose or tail the model was 1.7 feet from the wall at 30° which is just acceptable. Airship models with this degree of tunnel blockage are acceptable.
66 Airship Technology
The problem of the wrong Reynolds number remains: model tests were made at 6 x
106 when the full scale value was 2.1 x 108. The subsections in Airship Drag contain many references to the error in skin friction and form drag which wrong Reynolds number will cause. The problems arise due to the different development of the boundary layer. It has long been established that the most important objective is to ensure that the transition from laminar to turbulent boundary layer occurs at about the correct position on the model. This is done by ‘tripping’ the boundary layer usually with some added surface roughness. This will change the skin friction coefficient of the body but preserve the form drag and lift. In the subsection, The Bare Hull, it was shown that skin friction drag dominated the hull drag so drag results from roughened models may be in significant error. Gomes (1989a and 1990) went to considerable trouble to ensure that his tests, which have been quoted throughout this chapter, were
as realistic as possible. For pilot tests on the Skyship 600 model he covered the model with fishnet or fine mesh human stocking material. The fine mesh stocking suppressed early flow separations but, because it raised the skin friction coefficient, the drag exceeded a full scale value by 60%. Fine tuning the roughness would therefore be necessary to obtain accurate drag figures. For the YEZ-2A model Gomes decided that it was impractical to use stockings owing to the size of the model. He therefore obtained the required degree of surface roughness by spraying the polyurethane hull with a grey paint from a distance of approximately 60cms. This gave a coarse roughness "which was felt to closely reproduce the fine mesh stocking type of surface roughness" used for the Skyship 600 tests. Experimental Values x Upper surface + Lower surface
Model incidence = 10°
y Side
Lines are CFD values Hy 2° a —|,——
ae
io)=e
{ °o Le)
= a
|
r a a S i eex
= tan! {(V + vj cosb)/(QRx + vj sin >)} = tan!{(J + % )/(x + A tand)}
(5.20)
and,
where:
Figure 5.4.
Forces and angles of a blade section.
114
Airship Technology
All the variables in these equations are known except vj . For normal speeds vj « V and may be neglected in cruising flight. However, vehicles like the airship, this is not the case. For practical designs it neglect (vj sin o)/QR relative to x so simplifying Equations 5.16, 5.17
aircraft cruising for slow speed is accepta*le to and 20 to:
dL = “pcQ?R3( x2 + (J + 2 )*)a,(0 — o)dx
(5.2
dD = YapcQ?R3( x2 +(J + A )*)Cddx
(5.22)
o = tan! {(V + vj cosd)/QRx}
(5.23)
and,
The thrust T is found from Equation 5.18 by integrating over the length of the lifting blade and multiplying by the number of blades b, remembering that @ and c will most probably be functions of r, that a; and Cp are functions of incidence and Mach number and using a guessed value for vj which might have been determined by using Equation 5.10. The value of vj is then re-estimated for the new value of T, the final value of which is obtained by iteration. The accuracy of the final result largely depends on the precision with which vj is estimated. Improved estimation methods for vj are discussed next. Once the value of vj is found Equation 5.19 is used to determine the propeller torque Q. This is found by integrating dH.r over the length of the blade and including any contribution from the non-lifting root section and multiplying by the number of blades b. It will be noticed from Figure 5.4 that the induced velocity has a component of yj; sin in the direction opposite to the rotation of the propeller. This is the swirl velocity which represents a loss of propulsive momentum. More heavily loaded fans have inlet and/or outlet guide vanes to minimise this loss.
Propeller performance estimation charts exist which are used to relate thrust and power for a known design. A first approximation for propellers fitted to light aircraft is given by the ESDU data sheets mentioned in the list of references. Induced Velocity Estimation.
The simple actuator disc approximation may be improved by assuming that the
annulus swept by a blade element dr can be treated as an actuator disc producing a thrust b.dT. The equivalent of Equation 5.10 for the annulus at radius r is then: b.dT = 2p(2ar dr)(V + vj)vj
(5.24)
From Equations 5.21 and 5.24 vj can be determined. The resulting equation is nonis small, is made. For most linear unless a further simplification, namely that propellers this assumption cannot be justified. Well established mathematical
Propulsion techniques are used to solve the sum of the contributions way similar to that described Both the simple actuator
for vj from in the disc
115
for each spanwise annulus. The total thrust is then each annulus. The propeller torque is found in a last section. and the annulus model assume that the two-
dimensional lift is maintained out to the blade tip. This is not the case near the blade
tip where loss of lift occurs. A simple but moderately effective method of allowing for this loss is to assume that a blade produces no lift outside a radius of BR, where B is a number less than unity. The blade still produces drag in this region and therefore the integration for torque extends to the tip i.e to x = 1.0. Prandtl (1919) derived the
following expression:
B=1- VCrT/b where: Cy = T/pA(QR)?.
(5225) Typical values ofB are about 0.97.
Prandtl's analysis. which led to Equation 5.25, considered the flow generated by the lifting blades. The lift on the blade is equivalent to a circulation T about the quarter chord which is related to the local lift per unit length. This circulation is shed as vortices which trail behind the blade, each vortex moving under the influence of all other vortices. This includes the vortices from the other blades. Betz (1919) suggested that the induced power of a propeller was at a minimum when the vortex wakes from the blades move axially as if they were a solid screw surface. This hypothesis was later proved by Theodorsen (1948). Prandtl assumed that Betz's hypothesis was correct in which case the induced velocity across the disc would be constant. He further assumed that the curved sheets of the screw surface could be replaced by a series of two-dimensional sheets, based on the premise that the radius of curvature of the outer parts of the sheets is so large that they can be considered as doubly infinite straight strips. Assuming further that the blades have no drag then the element of thrust on an annulus of width dr at radius r is: dT = 2nrp(V + vj cosb)2vj cosd dr
(5.26)
and from the Kutta-Zhukowsky theorem: dT = dL cosh = pbWI cos@ dr
(3.27)
where W is the total incident velocity (Figure 5.4). Noting that W sing = V + vj cosd (Figure 5.4), eliminating dT between Equations 5.26 & 5.27 gives: vj = bI’/4ar sind
(5.28)
The complex potential of the flow produced when the series of two-dimensional sheets hypothesised above move relative to the surrounding air with velocity w is known to be:
116
Airship Technology
® + i ‘Y = (ws/n)arccos e%2/s
(5.29)
where © is velocity potential, Y the stream function, i = V(-1) and z = x + iy. the complex variable. The spacing between the vortex sheets, s, is given by: s = (2nr/b)sinod
Nn (5.30)
The required circulation to maintain the sheet arrangement is given by the difference between the velocity potential immediately above and immediately below the sheet. For simplicity, take the sheet on the x axis, for which ¥ = 0, the velocity potential difference a distance ‘a’ away from the edge gives the circulation I as:
I = wsk
(5.31)
k = (2/m) arccos e7%a/s
(5.32)
where:
k is called the circulation factor. If vj = 2w, then from Equations 5.30 & 5.31: vj = bI’/4ark sind
(5.33)
Comparing Equations 5.28 and 5.33 it is seen that representing the finite number of blades has introduced the factor k into the denominator of the expression for vj. Interpreting ‘a’ as the distance from the blade tip, R - r, then Equation 5.32 may be written as: -
k = (2/n) arccos e-f
(5.34)
f = “Ab(1-x)sing
(5.35)
where:
This relation is plotted in Figure 5.5. At the blade tip f = 0 and, as Equation 5.31 shows, I = 0. Increasing the number of blades leads to a larger value of f for a given x and , which means that the circulation is more nearly equal to the ideal value,
Goldstein (1929) solved the problem without making Prandtl's simplifying assumption regarding the vortex sheets. His analysis implied a certain variation of the inflow angle with radius which resulted in a more complex expression (actually a semi-infinite series of modified Bessel functions) for k than Equation 5.35. Graphs of
k against x for constant
¢ and a value of b may be found in his paper. Lock (1931)
extended the range of application by assuming that Goldstein's values of k applied
with reasonable accuracy to any practical propeller loading distribution so providing a general method of performance calculation. "
Propulsion
Figure 5.5.
Variation of circulation factor k with f.
Rolled up root vortex
Circulation distribution
Propeller blade
va
\
‘
Hub
Rolled up root vortex
P
Propeller disc Rolled up tip vortex
fle up tip vortex [_
Cd alice! Direction of vortex rotation
Figure 5.6.
1\7
_Diagrammatic representation of vortex roll-up behind a propeller blade.
118
Airship Technology
The above work may be criticised because it does not include the effect of the wake contraction shown in Figure 5.1. This omission means that the theory is only applicable to light propeller loadings. Theodorsen (1948) made allowance for this effect by starting from the final contracted wake and calculating the ideal propeller which generates it. Theodorsen (1969) extends the method to static propellers. These methods all assume that the shape of the vortex wake is fixed and is described by the parameter s. Such analyses are called, for obvious reasons, ‘fixed wake models’. The wake can also be conceived as a series of vortices springing from the trailing edge of a blade, the strength of each vortex being determined by the change in bound circulation each side of the point of origin. Each vortex will move under the influence of all other vortices with the exception of the bound vortices on the blade. Within a short distance behind the blade the vortices move towards the tip and the root region where they combine to form two concentrated vortices which trail behind and below the propeller. This is shown diagramatically in Figure 5.6. The root and tip vortices do not stop as shown in the left hand diagram but continue to contract and move downwards. For clarity only the first circuit has been shown. The bound circulation is approximated by a series of constant values as shown in Figure 5.7 and the trailing vortices are broken into straight segments. The velocity which any one of these vortex segments induces at a point P is given by the Biot and Savart law: dV =I cos B ds 4n r
(5.36)
where the symbols are defined in the inset in Figure 5.7. In this example dV at point P is normal to and points into the plane shown. The position of each segment is obtained by iteration. Methods to speed up the convergence have been developed, in
particular the choice of the initial positions of the vortex segments. For more details readers are advised to consult papers by Langrebe and his co-workers, for example Langrebe (1971). Propeller Performance A Comparison of Theory and Practice
Propeller performance is measured on a test stand, usually in the open air, for
static conditions. The main problem is obtaining very low wind conditions in which to make the measurements. Forward speed tests are made in a wind tunnel. The maximum size of propeller which may be tested is determined by the size of the
working section or the size of the airstream in a blower tunnel. For closed working
sections the walls must be sufficiently far away so that they do not affect the flow at
the propeller. For closed circuit wind-tunnels the energy put into the airstream must
be distributed throughout the full flow before it re-enters the working section. If not
the test is equivalent to a propeller operating in a distorted entry flow. If these conditions are satisfied the propeller area is restricted to less than 20% of the working section area.
Propulsion
Approximate circulation distribution
119
True circulation distribution
Blade
Concentrated tip vortex filament Discrete inboard vortex filament
= (2.8)
Velocity induced a vortex
Figure 5.7.
by
element
Vortex wake modelling behind a lifting blade.
Results from propeller prediction and test are usually presented in non-dimensional form. These are the thrust Ct and power coefficient Cp where: Cy =T/pn?d*,
Cp =P/pn?d4
Here n is the number of revolutions per second and d the propeller diameter. The efficiency n is defined by n = TV/P = CfTJ/Cp where J = V/nd =V/QR. The blade activity factor AF is a measure ofthe solidity which is the ratio of the total blade plan area to the disc area. It is defined by:
120
Airship Technology
Experimental Points
Blade Angle
|
Code
at x = 0.7 degs
0:24
0:18
Coefficient Thrust Cr
0:06
0
O04
O08
12
16
2
Advance Ratio J
Figure 5.8.
Full scale propeller wind tunnel test results C; vs. J.
Propulsion
AF
121
100000 i ex?dx 16 % d
Activity factors are normally in the range 100 - 150 for aircraft propellers. The pitch of a propeller, p, is the distance it would advance in one revolution if there were no slip. From Figure 5.4 it may be seen that p = 2nR .tan®. For a constant pitch propeller of diameter D the section pitch angles are calculated from: 0 = tan! {(p/D)/1x }
A typical comparison of theoretical and experimental results is given in Figures 5.8 & 5.9. These results are quoted from data supplied by Paul Methuen of Dowty Aerospace Propellers. Figure 5.8 shows the comparison of strip theory with experimental result obtained in the RAE 24 foot open jet wind tunnel with a full scale 2.77 m four-bladed propeller. The curves are for different blade pitch settings defined by the value at 0.7 radius denoted as 0,;, The results shown are for the thrust coefficient Cy against the advance ratio J. The strip theory curves were derived by calculating the thrust at measured powers.
Experimental results:
J=0.67
x,
J=0.76
0,
J=0.89
z
Strip theory results shown
1600
: Indicates pressure tapping stations
|
C, local Chord Paremeter Vheiicai. Loading =. Non-Dimensional
Figure 5.9.
Span
'% scale four bladed pressure tapped propeller measurements.
122
Airship Technology
Agreement is good except at high values of Cy, at lower values of J and for the
higher values of 0 7. This is the region of the performance map where stall] will occur and work is in hand to improve predictions of three-dimensional stalling characteristics. Figure 5.9 shows the results obtained in the Southampton University 11 x 8 foot wind tunnel from a 0.69m diameter four-bladed propeller. From chordwise pressure measurements made at six spanwise locations the loading parameter (lift) has been derived for three advance ratios. J was changed by varying the rotational speed of the propeller at a fixed tunnel speed of 49ms-!. The loading parameter has also been estimated by strip theory and this is shown by the continuous curves. The comments on agreement made on Figure 5.8 above also apply here.
Ducted Propellers.
Consider the open propeller represented in Figure 5.1 and enclose it in a duct as shown in Figure 5.10.
Plane
A
BiaeG
D
Open propeller
Open propeller streamtube
Total Pressure
: ee Static
P-Pool-
ond
Propeller
Ducted
—
Velocity
V
Figure 5.10. Comparison of an open with a ducted propeller.
Open
Propulsion
123
The duct is chosen so that upstream at plane A and downstream at plane D the flows are identical. Both propellers will produce a pressure rise Ap across their respective discs. Thus the resultant thrust of both systems must be the same since the rate of change of momentum is the same. If Q is the mass flow at any plane of either system then the total thrust T is given by:
T =Q((V + v2) - V) = Qv,
(5.37)
The thrusts produced at the propeller, Tp, and the ducted propeller Tp
are:
Tp= Ap mR?
(3:38)
Tp = Ap m(R + AR)?
(5.39)
and,
respectively.
However Tp must equal T but Tp must be combined with any thrust
which is produced by the duct. Subtracting Equation 5.38 from 5.39 gives the thrust produced by the duct, in this case in a negative direction, as:
Tduct = -Ap2nR. AR = -Tp 2 AR/R
(5.40)
No net circulation
Velocity ii
V : Propeller plane ————
Shed
vortex sheet
With circulation
Figure 5.11.
Effect of simple duct on flow about a propeller.
124
Airship Technology
Although the pressure rise across the two discs is identical, the velocity through the two discs will not be the same, the velocity through the larger ducted disc being the lower. Bernoulli's equation between planes A and B shows that the static pressure at plane B is higher for the ducted case. The comparative pressures and velocities for the two flows are also shown in Figure 5.10. Figure 5.11 shows the way in which the force is generated on the duct. The upper
half of the diagram shows a propeller and two streamlines (solid curves). A simple cylindrical duct is inserted and this alters the streamlines (dashed curves). The flow adjusts around the leading and trailing edges where equal and opposite suctions are produced. There is no net circulation around the duct. However in practice the situation is that shown in the lower half of Figure 5.11. The Kutta condition must be satisfied at the trailing edge and the leading edge of the duct must be rounded to avoid a velocity discontinuity. Thus the surface at the leading edge now supports a suction due to the accelerating flow around it so giving rise to a thrust force while the flow from the trailing edge will result in a time-dependent and possibly periodic shedding of vorticity. This is the boundary of a jet. In the case when the pressure at the outlet plane is equal to poo a steady vortex sheet will be formed. The development of circulation suggests that the duct may be modelled as a distribution of ring vortices along its camber line. The strength of the vortex rings is chosen so that the resultant flow from the free stream, the propeller and the vortices is tangential to the duct surface at a number of points equal to the number of vortex rings. The simplest model is by Weissinger who used a single vortex ring situated at the quarter chord of the duct. The strength of this vortex is determined by making the resultant flow tangential to the duct at the three-quarter chord position. The geometry of the model is shown in Figure 5.12.
Figure 5.12.
Approximation to a ring aerofoil.
Propulsion
125
The induced velocity v; is found from:
ig ea
MDipa
; Ca D
(5.41)
(Piya Piya
where the function f arises from the integration round the vortex ring to give the velocity at any point on'the three-quarter chord circumference. Values of the function f, taken from KUchemann and Weber (1953), are shown in Figure 5.13. To complete
this calculation it is necessary to determine propeller flow as a function of axial distance.
Figure 5.13.
the radial velocity variation of the
Variation of ‘f with ring aerofoil parameters.
126
Airship Technology
This is most simply done by approximating the propeller wake as a single tip vortex helix the geometry of which is shown in Figure 5.14. Let p be the pitch of the helix with generating angle 0 and the differential element of the vortex ds be located by the radius vector R from the origin. For the point Zp, the radius vector r to the vortex element is given by: r=iRcos9
+ jRsin® + k(p0/2n - zp)
(5.42)
where i, j and k are the unit orthogonal vectors. So ds = (-iR sin 9 + jRcos®@ + kp/27)d0 as it is orthogonal to R. Using the Biot and Savart law, the axial velocity v(Z) induced at Zp for a vortex of strength y is:
v(z)
tyes = ea
ER, r =o
2 ps 1 +z, /(R?4z,”)
(5.43)
The variation of v(z) may most easily be expressed as:
v(z)= 1+
Vi
_dR
(5.44)
V(1 + (/R))
where vj; is the induced velocity at the propeller disc, z = 0.
(a)
Figure 5.14.
Geometry of a helical vortex.
(b)
Propulsion
127
Using the continuity condition: (V + vi)nR? = (V + v(z))nR?
(5.45)
where Rz is the radius of the streamtube at z. The slope of the streamtube is given by: dR; /dz = -R(dv(z)/dz)/2(V + v(z))
(5.46)
and the radial velocity vjp is:
ViR = (V + v(z)) dR; /dz
(3.47)
Combining Equations 5.44, 5.46 and 5.47, with Equation 5.10 providing the value of vj, the information is available to find the thrust of the duct using Weissinger's approximation. If © is the tangent to the mean camber line at 3c/4 then:
(MIR =Vi Vt
Vi)
i=) O
(5.48)
where R(3/4) and v(3/4) are the values of the variables at the 3c/4 plane. Equations 5.41 and 5.48 allow vj; to be eliminated and I determined from:
P=
{tDyyal £(c/Dig D345 /Dya )}[-ViR(s/a) - O (V + V(3/4))]
AD)
The duct thrust T, is then:
Tp = -PViR(1/4) PaDyj
(5.50)
The duct can increase thrust and thrust/power ratio particularly at zero forward speed, but its drag is not negligible. This is particularly true if a duct is designed to enhance static thrust. The drag of the duct may be estimated in the standard manner. An indication of the accuracy of this estimation method and the effect of duct drag is shown in Figure 5.15 which is taken from McCormick (1967). The duct shape and position of the propeller is also shown in that Figure. The continuous lines show the estimated performance using both the inner duct surface contour and the mean camber
line to determine the duct circulation I’. The abscissa of the graph is a form of thrust coefficient based on the total thrust T. The use of the inner duct contour is much superior. The inclusion of duct drag is of greatest importance for low thrust conditions. For higher thrust coefficients an underestimate of the thrust ratio results. However, agreement is reasonable bearing in mind the simplicity of the mathematical model. It suggests that a more representative model using sources for thickness effect as well as a distribution of vortices will give better agreement. Such a model will only be correct if the flow stays attached to the duct both internally and externally: failure to do so will probably result in a poorer performance than that of the isolated propeller.
128
Airship Technology
2 0
16
o x
c a @
pea b/
42
=
:
ao
|
Based on inner surface (C, = 0.2)
Based on inner surface (C4 = 0)
O/
oO
x
=
+
Based on mean line (Cg = 0)
3 oe
é
Experimental Values
2
Rotor Pitch/ | Symbol
Gre
Diameter
0-4
0
e
0
3°2
6-4
96
12:8
Ti(p V? D*)
Chord line
Plane of propeller Flow direction —————7
Axis
of revolution
Duct Geometry
Figure 5.15.
Ducted propeller performance.
16
Propulsion
129
Ducting a propeller may also have operational advantages by offering protection to people when the airship is on the ground, offering some protection to the propeller from trailing lines and ground vehicles and allowing the installation of acoustic liners to reduce external propeller noise. For heavily loaded propellers the installation of inlet, and possibly outlet, guide vanes will reduce swirl and enhance the performance.
THE PRIME MOVER The extraction of energy from added fuel and its conversion to a form which can do useful work is described by a number of thermodynamic cycles. Three cycles dominate currently used prime movers. These are the Otto cycle used by the reciprocating, spark ignition internal combustion engine (abbreviated to ICE). the Diesel cycle of the compression ignition ICE and the Joule (in the UK) or the Brayton (in the USA) cycle for the gas turbine ICE. The basic cycles are not described here but may be found in all basic propulsion texts. The choice of the prime mover is influenced by the overall cost. This includes the initial cost, the running cost which incorporates mainteriance expenses and fuel as well as the effect of the weight on the payload which can be carried. Choice also depends on the availability of an engine in the correct power range unless the project is able to bear the cost of the development of a dedicated unit. In this chapter the discussion will concentrate on the basic characteristics of engines which use the three principal cycles. The variation of shaft speed and specific fuel consumption (SFC) with power for the three cycles is shown in Figure 5.16. Comparative Performance
The relation between power and shaft speed is a discrete curve as shown in Figure 5.16. The specific fuel consumption as a function of power depends on the size of the engine, the larger units achieving a thermal efficiency of 30% but the smaller
engines only reach 20%. The SFC is shown as a band in each case. The maximum power does not coincide with lowest specific fuel consumption in the case of the Otto and Diesel cycles. The SFC of the three cycles is in order of improving performance, the Joule, the Otto and the Diesel. Figure 5.17 shows the variation of engine specific
weight with shaft horsepower as a shaded band for shaft turbine engines. There are insufficient modern Diesel and Otto cycle aircraft engines in production to produce similar curves for them so some examples are shown as discrete points. The Joule
cycle gives the lightest and the Diesel the heaviest engine. It is essential to consider
the combined engine and fuel used weight in typical missions to decide the preferred engine type(s).
130
Airship Technology
100
80 ie a
;
=
.
~
2
60
a
oH
ee Ww
Ee
eal
20
Otto Cycle ICE
i
@)
0
c Qa
re
~~
2
O
WL.
% RPM Max. RPM Max.
”
te Qa
c
en
2
O
u.
(v9)
%Max. RPM Diesel Cycle ICE
0
0 0
20
40
60
% of Maximum
80
100
Power
Figure 5.16. Variation of specific fuel consumption (SFC) with power for the main types of ICE.
Propulsion
131
The missions which are normally cited as being especially suitable for airship operation take advantage of the excellent loiter characteristics of buoyant lift. Various authors have discussed these roles and readers are referred to the papers presented at the 18th Annual Symposium of the Airship Association and reported in ‘Airship’ No. 98 for a balanced
discussion.
In one
of those papers
Hillsdon
(1992)
succinctly
summarised the conflicting factors when choosing an airship propulsion system. The following is a quotation from that paper.
"The endurance of an airship is governed by the difference between permitted heaviness at take-off and permitted lightness at landing. The only way out of this constraint is reballasting during flight. Reballasting in flight involves either pick-up from the surface (possibly incompatible with many surveillance roles) or recovery of ballast water from engine exhaust. The latter course is extremely difficult with gas turbine engines and is deleterious to fuel consumption with any engine. At low flight speeds, power for propulsion may well be exceeded by the power loading of the surveillance mission equipment and essential services. The engine efficiency of installed systems is thus paramount for the airship. An overall energy balance must be computed for all the mission profiles projected for the airship and used together with the appropriate spectrum of winds for the operating areas to evolve an overall power spectrum for the ship."
T
ss
D- Diesel Engines P- Otto Cycle Engines =
| | |
0-4 kg/SHP Weight Specific
Engine
Figure 5.17.
Shaft
Horsepower
(thousands)
Engine specific weight vs. shaft horsepower.
132
Airship Technology
These requirements determine the power units which meet the basic power demands of the missions. The third point above states this will be determined by the speed requirements and the provision of services for the operation of the craft and its payload. Thus electricity generation, provision of hydraulic power etc. required to power ballonet inflation fans, control surface actuators, cabin conditioning as well as surveillance radars, winches etc. for the different roles have all to be assessed for their
part in determining the base load. This power must be provided most efficiently. For the larger airships part of the base load will be provided by an auxiliary power unit which is run at a constant speed to give optimum fuel efficiency.
CODAG = Combined
Airship volume 100,000m"
Diesel & Gas Turbine
Cruise height=5,000 feet
2 x CRM BR - 1/200
ASI Temperature 20°
Diesels + GE Gas Turbine
______—
All Gas Turbines GE T700-405
“Zo NZ ra
) Engine (Ibs. Weight Fuel Plus
0
0:8
16
242
"92
Days on Station
Figure 5.18. Endurance comparison of aCODAG and an all gas turbine solution.
40
Propulsion
133
The problems following an engine failure must be considered against the role requirements. If it is essential that the mission continues after the loss of an engine then either the remaining engine(s) can be run at higher power settings or a spare engine can be started and its output added to replace that of the failed unit. In the latter case the mission can continue as planned but all operations will pay the penalty of having to carry the weight of the ‘spare’ power-plant as a loss of payload. In the case of the YEZ-2A this solution was adopted with the stern turbo-prop unit only being used for specific phases of the flight, like a high speed dash, or an emergency. The choice of the lighter, but slightly thirstier, turbo-prop unit for this role compared with the Diesel main engines demonstrates the compromise which the designer has to make. Such a choice carries with it the need to provide technical and maintenance support for both types of engine as well as two types of fuel supply unless the gas turbine is adapted to burn diesel fuel or the diesel JP4. The benefits of the YEZ-2A CODAG (COmbined Diesel And Gas turbine) powerplant installation relative to an all gas turbine installation can be demonstrated by calculating the endurance of two identical airships apart from their power plants. Figure 5.18, copied from an article by Hillsdon (1986), shows the time at which the better fuel consumption of the CODAG system overcomes its heavier installation weight when compared to an all gas turbine drive system. The endurance varies with the operational airspeed required, the higher the speed the greater the power needed and, therefore, the sooner the CODAG system becomes superior. The choice of the CODAG installation for the long endurance missions intended for the YEZ-2A is demonstrated. Figure 5.18, however, also shows that for short range passenger and freight operations, for which the higher airspeed but relatively short endurance is likely to be the norm, the all gas turbine installation will probably be preferred. The larger the lifting capacity of the airship the greater is the choice of propulsion arrangements. The advantage, however, of keeping the number of engines and propulsors to a minimum commensurate with operational requirements - including safety - is no different from that for other types of airborne vehicles. The arrangement and choice of the propulsion system can play an important role in airship control and may improve the overall aerodynamic performance. This is considered in Chapter 3.
ENERGY SOURCES Since lift from buoyancy is constant at a fixed temperature and pressure, an airship will climb as fuel is consumed unless other means are used to trim the craft. The Graf Zeppelin was fuelled with a mixture of propane and hydrogen which has a
density very similar to air. As the fuel was consumed it was replaced by air so keeping the craft weight approximately constant.
134
Airship Technology
The characteristics of various fuels which have been considered suitable for aviation use are shown in Figure 5.19. The points show the volume and the weight required to produce a unit heat release relative to that for kerosene. Clearly fuels which lie to the left and below the o, which marks kerosene itself, are desirable. Those shown in the
diagram are boron based from which toxic by-products result so for general use they may be discounted. Hydrogen stands out as being of special interest due to its low weight per unit heat release however it must be remembered that the container for
hydrogen is expensive, heavy and large relative to a normal aviation fuel tank. No concept design of a hydrogen fuelled airship relative to a conventional layout exists. To gain some indication of the gains and losses, results from a Lockheed study which compared a 1011 airliner using JP4 with a variant using hydrogen fuel to perform the same mission are given in Table 5.1. The figures show that for the hydrogen burning aircraft the take-off weight and that of the fuel is reduced, but that the fuel volume is
increased. This reduces the lift/drag ratio significantly. It is however concluded in the study that hydrogen fuel would be an acceptable alternative to oil based fuels if the latter became in short supply. On this basis one would expect hydrogen fuel to be a
useful source of energy for airships.
O = Kerosene
_ Ethyl Alcohol
Petrol. 4, ‘Methyl Alcohol Pentynes
a8 raffins
Acetylene—-
. Methane
,Decarborane v4«Pentaborane
Unit Release Kerosene /w.r.t. Weight Heat
)
1
2
3
4
5
Volume/Unit Heat Release w.r.t. Kerosene
Figure 5.19. Weight and volume per unit heat release w.r.t. kerosene.
Propulsion
135
If a safe method of containing it in gaseous form could be found, its use to increase buoyancy in particular phases of the flight would be a further bonus for the airship. Recent, but unconfirmed, reports suggest that a Russian airship prototype, the ALA600 developed by the Moscow Aviation Institute, is to be fuelled with hydrogen which will both provide buoyancy and be burned in the turbo-prop engines. For the immediate future, however, conventional hydrocarbon fuels appear likely to be the norm.
PROPULSION, PERFORMANCE AND CONTROL The quotation in the section on Comparative Performance relating the influence of heaviness to the choice of the propulsion system did not include in the potential solutions the use of engine generated thrust. The Airship Industries series of airships have demonstrated that vectoring the thrust of the propulsion engines improves both the performance and the control of the craft. In those airships the pitch of the ducted propellers is reversible so enabling relatively rapid changes in thrust. In addition the ducted propeller can be rotated from 90° up to 120° down so enabling the thrust line to be directed for both performance and control reasons. In a vectored thrust design the mechanical arrangement of engine(s) and propeller falls into one of two basic classes. The engine can drive the propeller directly in which case it rotates as the thrust is vectored. The engine lubrication and fuel supply must then be designed to operate for extended periods in an inclined position. Most engines are not designed for such extended operation and the cost of the modification, when only a small number of engines are required, is high.
Table 5.1. Comparison of a conventional with a hydrogen fuelled Lockheed 1011 transport aircraft
Fo-tnen we|28am [215350 | 0 Rrebesee frosting ip= scm | om | eae |
136
Airship Technology
The inertia of the rotatable propulsion unit is significantly increased so requiring larger control actuators. Fuel and engine control signals have to be provided through a rotating joint. The alternative is to mount the engine(s) in the non-rotating frame of reference and transmit the drive through a right-angled gearbox to the propeller with a consequent weight, performance and maintenance penalty. The trend is to accept the penalty of a right-angled gearbox and mount the engines on or in the gondola of non-rigid airships. In symmetrical propeller arrangements, as in the Skyship airships, this allows the designer the further option of interconnecting the propellers with a transverse shaft on to which all engines drive through freewheels and/or clutches so enabling both propellers to produce thrust in the event of an engine failure. A method of increasing the deflection of the propeller slipstream, so reducing the mechanical tilt needed, has been adapted from tilt-wing experience. Nayler (1988)
describes the method proposed by Advanced Airship Corporation for the ANR airship which was under construction in the Isle of Man. The stub wing attached at the rear of the gondola, which has a propeller and engine at each wing tip, can be rotated from 759 up to 30° down. The deflection of the slipstream may be increased to an equivalent of 90° by the deflection of a single trailing edge flap. Beyond the end of the wing the slipstream deflection due to the trailing edge flap is much reduced. The non-rigid airship design restricts the mounting of the propulsors to the gondola since this is the only structure which can readily dissipate the thrust throughout the whole craft. The designers’ scope for influencing the aerodynamics of the whole vehicle by the air flow through the propulsors is therefore limited. The rigid or semi-rigid design has more flexibility, the range of possible mounting points being illustrated in photographs of the Zeppelin and US designs. Positioning a propulsor at the tip of the tail cone has two advantages. First it ingests the boundary layer which has developed along the hull. Equation 5.12 above shows that as V. the incident velocity on the propulsor, decreases the power required per unit thrust decreases. (The thickness of and momentum loss in the boundary layer is discussed in Chapter 3). Secondly the propulsor entrains the flow around the hull which accelerates the boundary layer, thus reducing the risk of flow separation. This reduces
the profile drag of the hull but will have little effect on the more important skin friction drag. By itself this gain is unlikely to justify the complication, weight and
moment of inertia increase which will result. For further discussion of the ingestion of boundary layers into a power-plant readers are referred to Ktichemann (1978), page 315, The tail mounting of a propulsor can, however, provide considerable control advantages if its thrust is vectored. This may be done by moving the propeller axis of rotation to provide pitch and/or yaw control. Such a system is described by Kollmann (1992) for the projected Zeppelin NT. Alternatively, control surfaces (rudders and/or elevators) mounted
behind a fixed axis propeller may be inclined to deflect the
slipstream. While this is mechanically more simple than swivelling the propeller its
efficiency is lower due to the drag of the surfaces as well as their weight penalty.
There has been much talk of copying ships and installing bow and/or stern thrusters for low speed control. Advanced Airship Corporation's ANR airship was planned with
Propulsion
137
both bow and stern thrusters. These were low velocity cold air jets which could be vectored to provide pitch and/or yawing moments and vertical and/or sideways forces. Actuator theory has shown that to maintain an acceptable efficiency such jets must
obtain their rate of change of momentum by using a large mass flow exhausting at a low velocity. Nayler(1988) described the system as follows.
"The low-velocity jet produces a minimum thrust of 27kg. and a volume flow of 160 m/min. Ballonet air is used as the supply reservoir boosted by two fans The air passes to the plenum chamber and thence to one pitch and two yaw ducts. With the Jans running stalled, thrust is immediately available when a duct valve is opened. This system will assist control at low airspeed (< 10 knots) and during the hover." Such a system avoids the complication of placing a fan at the thruster which could be difficult mechanically. It also provides a large reservoir of pressurised air which is maintained by the fans which each have the same maximum volume flow as the jets. The reduction of drag by the use of controlled blowing or suction has been actively researched for many years but has not yet found a practical application. One approach of considerable interest to the airship designer is to delay the boundary layer transition region so reducing skin friction drag. For this application suction is used to remove most of the boundary layer. Suction may be applied beneath a porous skin. Varying the porosity of the skin to reflect the potential growth of the boundary layer over a particular area allows a constant suction pressure to be applied. A major problem is the small size of the pores which makes them very vulnerable to becoming blocked by insects or general atmospheric dirt. An alternative approach is to use slots across the flow direction into which the boundary layer, which has developed from the upstream
slot, is sucked.
This
solution,
while
more
tolerant
of dirt, brings
mechanical problems. The maximum suction pressure is 1 bar. Large ducts are needed to remove the ‘dead’ air which in the case of the airship represents a buoyancy penalty. For both systems vacuum pumps are required and these must be powered from the engines. Injection of momentum into the boundary layer by blowing through a surface slot will not produce a laminar boundary layer. It is effective in delaying separation by reenergising the boundary layer. Since high pressure air may be used up to the slots the internal ducts are smaller than for the suction system. High pressure air is available from gas turbines, up to 10% of the compressor mass flow may be bled away. While this reduces the power output of the engine it does avoid the installation of more engine accessories. Suppressing separation allows the designer to use a hull shape which has a higher thickness ratio and is structurally better. Paki and Pipitone (1975) have made
calculations and wind tunnel experiments on such a shape using a radial suction slot
at the rear. Figure 5.20 shows the shape of their test body as well as the velocity profile. They have achieved accelerating flow almost up to the position of the slot so reducing the growth of the boundary layer. The slot suction permits the freestream flow to turn the sharp corner at the junction with and to remain attached over the short tail cone. A simple design exercise on an airship with a volume of 106 cubic feet
138
Airship Technology
operating suggested an endurance improvement of between 20 and 40% at most through slot the through in taken flow speeds. One interesting idea was to exhaust the control. speed low a swivelling nozzle at the tip of the tail cone so providing enhanced While an interesting concept, it requires further detailed study and some flight experience to confirm the advantages claimed for boundary layer control.
Transition Region
Turbulent boundary layer region Under favourable pressure gradient
Laminar boundary Layer region
Ratio Velocity Local
Non-Dimensional Length
Figure 5.20. Velocity ditribution over BLC airship body shape.
Propulsion
139
REFERENCES Betz, A. (1919). Schraubenpropellers mit geringstem Energieverlust. Keg. Ges. Wiss. Nachr. Math.-Physik. ESDU Data Sheets. 1. Estimation of uninstalled thrust of a constant pitch propeller. Data Sheet 83001. 2. Estimation of uninstalled thrust of a fixed pitch propeller. Data Sheet 83028. Engineering & Science Data Unit International Plc. Goldstein, S.(1929). On the Vortex Theory of Screw Propellers, Proc. Roy. Soc., A123 p440. Hillsdon, R. H.(1986). The Design Challenge of aLong Endurance Airship. Airship No. 74
Hillsdon, R. H.(1992). An Analysis of Viable Roles for Airships in the 21st. Century. Airship No. 98 Kollmann, H. (1992). Zeppelin NT - A New Development in Riogid Airship Technology. Airship no. 98. Kiichemann, D. and Weber, J.(1953). Aerodynamics of Propulsion. McGraw-Hill Book Company, New York. Kiichemann, D. (1978). The Aerodynamic Design of Aircraft. Pergamon Press. Langrebe, A.J. (1971). Analytical and Experimental Investigation of Helicopter Rotor Hover Performance and Wake Geomeiry. USAAMRDL Tech. Rep. 71-24. Lock, C. N. H.(1931). The Application of Goldstein's Theory to the Practical Design
of Airscrews. British Aero. Res. Council R&M 1377. McCormick, B. W.(1967). Aerodynamics of V/STOL Flight. Academic Press. Nayler, A.W.L.(1988). Advanced Airship Corporation: The ANR Described. Airship NO. 79 Paki, F.A. & Pipitone, S.J.(1975). BLC for Airships. Proceedings of Interagency Workshop on LTA Vehicles. Edited by J. F. Vitteck. Prandtl, L. (1919). Appendix to Betz, A., (q.v.) Theodorsen, T.(1948). Theory of Propellers , McGraw-Hill, New York. Theodorsen, T.(1969). Theory of Static Propellers & Helicopter Rotors, Proc. 25th. Annual Forum of the American Helicopter Society.
6 Materials P. Bradley
INTRODUCTION Many of the major material developments of this century, particularly in the areas of specific strength. have been instigated by the aerospace industries. These developments/advances have fundamentally changed the design of aircraft. The introduction of such materials as lightweight strong alloys, fibre reinforced composites or honeycomb materials have all had major impacts on the design of aircraft structures while improvements in high temperature properties through alloy development and manufacturing methods (e.g. directionally cast, single crystal, or metal matrix composite turbine blades) have transformed engine technology. Most of these developments, and especially the introduction of stronger lighter materials have helped airship design through lighter gondolas, nose cones, battens and tail fins. These consequences, however, are a spin-off from material developments aimed at conventional airframe rather than as airship specific developments. However. it is major developments in textile engineering that have been uniquely responsible for advances in airship design, and for this reason this chapter will consider those developments only. The advances in other structural materials have been adequately covered in other publications. Material development is a continuing process and no publication can hope to be more than a statement of the current state of the art. This chapter will attempt to look at the progress made in developing textile materials up to the current situation. and give some insight into the properties of textiles that are important in the designing and
building of modern airship envelopes.
DESIRABLE PROPERTIES
FOR AIRSHIP TEXTILE MATERIALS
Materials ideal for use in pressurised non-rigid airship envelopes would have the following properties:
142
Airship Technology Envelope Materials
e
High strength. The strength of the material possible size of the envelope.
will determine
the maximum
e
High strength to weight ratio to minimise the weight of the envelope.
e
Resistance to environmental degradation through temperature, humidity and ultraviolet light. These factors will determine the life of the envelope and its maintenance requirements.
e
High tear resistance to give damage tolerance.
e
Low permeability to minimise helium loss. Helium operational capability and increased operational costs.
e
Joining techniques that produce strong and reliable joints, not subject to creep rupture.
e
Low creep to ensure that the envelope shape is maintained throughout its life.
loss results in loss of
Ballonet Materials
e
Low permeability to both air and helium to minimise helium contamination or loss.
¢
Good flexure and abrasion resistance with no increase in permeability.
e
= Low weight.
DEVELOPMENT
OF TEXTILES FOR AIRSHIP USE
With the development of synthetic materials, the textiles used in airship
manufacture have undergone many changes. The development of these materials from those made of natural fibres through to today's totally synthetic materials is outlined below. Envelope Materials from Natural fibres
The early envelope materials were exclusively plied rubber coated fabrics, two
or three plies of cotton fabric having been bonded together using natural rubber. The cotton plies were biased relative to each other in order, it was claimed, to improve shear strength and stiffness and tear resistance. The natural rubber sealed the fabric
Materials
143
although permeability was controlled to a limited extent by altering the amount of rubber in the inter-ply layers, a ‘gastight’ fabric having more rubber and being heavier
than an ‘airtight’ fabric. To reduce permeability further, the envelope was ‘sealed’ with paraffin on the inner surface and both inner and outer surfaces of the ballonet. The envelope material was joined by a mix of sewn and bonded joints although the sewing was primarily to maintain good contact in the joint while the bonding agent. rubber cement, cured. However, under the influence of solar radiation. temperatures sufficient to soften the rubber cement were encountered and the rows of stitching were necessary to prevent joins from slipping apart, until cooler conditions prevailed and the rubber cement rehardened. The material had poor properties by modern standards. Houmard (1986) refers to a 650g/m* (19 oz/sq yd), three ply cotton fabric having a strength of about 0.7 kKN/50mm (80 Ib/inch). Because its susceptibility to weathering resulted in a rapid
deterioration in strength, not only was it necessary to use large safety factors but the envelope had a short life. Consequently the limited strength of natural fibre fabrics limited the size of non-rigid envelopes to somewhere about 8,500m’ (0.3 million ft’). Introduction of Synthetic Material
The replacement of natural rubber coatings by neoprene (chloroprene) rubbers was the first use of synthetic products in envelope materials. Although neoprene has a higher specific gravity than natural rubber, it has much improved weatherability and a lower permeability to gases. The envelope materials would be expected to require somewhat less maintenance and have a longer life, but the material still had a very limited strength. However, the introduction of base fabrics made from synthetic fibres gave large increases in strength for less weight. Polyamides (nylon) and polyesters (particularly Du Pont's Dacron) were two very early synthetics that were produced as filament yarns from which fabrics could be woven. The comparison with cotton is shown in
Table 6.1 below:
Table 6.1.
Properties of three fabrics
Tensile Strength
0.4-0.55
0.8
1.0
GPa (Ib/sq. in)
(60-80,000)
(120,000)
(140,000)
5
12
(720,000)
(1,700,000)
Tensile Modulus
GPa (Ib/sq. in)
144
Airship Technology
The potential for stronger, lighter materials is evident. The greater tensile modulus of the polyester coupled with its better resistance to hydrolysis (nylon can absorb approx. 4% by weight of water) made it the preferred yarn for the airship envelope materials. Polyester was the base textile of the lightest envelope fabric used in the Goodyear GZ 20, a 370g/m’ (10.9 oz/sq. yd), two ply, neoprene coated polyester fabric with a minimum specified breaking strength of 145kKN/50mm (165 Ib./inch). The largest nonrigid airship ever built, the 42500m? (1.5 million ft?) ZPG 3W, had an envelope constructed from two ply, neoprene coated, polyester fabric with a minimum breaking strength of 2.8KN/50mm (320 Ib./inch) in the warp direction at a maximum weight of
560g/m’ (16.5 oz /sq. yd). These materials gave much improved strength at lower weights than the natural materials, but still had only moderate weatherability and therefore needed regular envelope maintenance and frequent repainting. The method of making bonded joints was unchanged, and the joints remained prone to degradation. A further step on this path was the use by Airship Industries, in the Skyship 500 and 600 series ofa single ply polyurethane coated fabric. The base fabric. as before. was a plain polyester 550 dtex, 15 x 15 weave spread-coated on both sides with a polyether grade polyurethane and with a layer of PVDC Copolymer (Du Pont's Saran) bonded to one (inner) surface to reduce helium permeability. The material gave a breaking strength of 1.85 kN/50 mm (210 lb/in) for a weight of 370-400 g/m” (10.9-11.8 oz per sq yd). The polyurethane coating gave better weatherability and therefore reduced the need for envelope maintenance with less frequent painting required although this was only true of temperate climates as in the hotter, humid tropics, the degradation suffered by the polyurethane is accelerated. The method of making the joints was modified in these materials. Instead of an overlapped bonded and sewn joint, the material was butt joined with the load carried by butt straps, strips of envelope material bonded across the join on both the inner and outer surfaces. No joints were sewn. The straps on each surface were of different widths so that the change in stiffness across the joint was gradual and did not induce stress concentrations. The polyurethane adhesives were very effective in carrying the load in shear such that in strength tests, it was usual for the material to fail rather than the joint pull apart. The adhesive maintained its strength to higher temperatures also, loss of strength being seen only at temperatures in excess of 75°C. Additionally. the bonded joints did demonstrate a much better resistance to environmental degradation. One other feature of this material is that it has allowed the construction of the envelope to be changed. Previously each gore from which an envelope was constructed had been made from a large number of lengths joined at their edges, giving a large number of joins perpendicular to the axis of the gore. The reason is said to be that the material was stronger along the roll length than across the roll width (the terms warp and weft not being applicable to two or more ply materials) so that the
stronger direction was oriented in the more highly stressed circumferential direction of the envelope. The material for the Skyship 500 and 600 envelopes. being a single
ply. had essentially the same strength in the warp and weft directions but with a
Materials
145
greater extension in the warp. Cutting the gore from the length ofthe roll gave several
benefits: the greater extension along the warp was belicved to help relieve stress concentrations introduced during the manufacturing process; the reduction in the number of joints required to manufacture the envelope reduced the manufacturing time, reduced the length of join and thus the risk of poorly bonded joins and reduced the envelope weight. Laminate Materials
The most modern materials which have been developed as envelope materials have been laminate materials. The feature of laminates is that overall properties of the laminate can be tailored by the judicious selection of appropriate components of the laminate. Thus the laminate will consist of components for load bearing, gas retention and to protect against weathering and the environment, bonded together with an adhesive that allows for ease both of laminate manufacture and ofjoining the material. A typical laminate is shown in Figure 6.1. Outer surface
Environmental / weathering protection layer
Adhesive layer Gas retention layer
Adhesive layer _
Woven load bearing layer
Adhesive layer Inner surface
Figure 6.1. |Section through typical laminate material.
The structure has an obvious inner and outer face, with the gas retention component sandwiched between the load bearing and weather resistant components. This allows for joints to be made by bonding directly to the load bearing component.
146
Airship Technology Load Bearing Component
The requirement for an envelope material to be flexible mandates that the load bearing component will be a woven fabric. A large selection of synthetic fibres are now available of which a number of potential suitable candidates are given in Table 6.2 together with relevant properties.
The largest non-rigid airship ever built, the 42500m’ (1.5 million ft?) ZPG-3W, had an envelope whose load bearing component was two plies Thirty-five years on, polyester still remains the preferred fabric has a good combination of required properties, high strength and hydrolytic stabilities. When fabricated, typically, from a
of a polyester fabric. because, on balance, it and good dimensional low twist 1000-denier
yarn, a 13 x 13 count with low crimp and a balanced weave, the cloth has minimum
elongation, low crimp interchange, and a high tear resistance. Polyester also has the benefit of a long provenance. The competing fibres, while having excellent properties in many respects, do not currently achieve such a desirable combination of properties and in some instances have a significant drawback. This may change as greater experience of the newer fibres is gained and their less desirable properties, in this context, are improved or can be compensated for by some other means. Polyamides have a very similar specific strength to polyester but have greater extensibility and failure strain with a lower modulus; they are also subject to hydrolytic degradation. A polyamide envelope would have poorer dimensional stability with probably a shorter life. Aramids have a much higher specific strength, a greater modulus and a much lower failure strain, being originally developed as a cheaper alternative to carbon fibres as reinforcement for resins. Superficially they could significantly reduce the weight of an envelope taking only strength into account, and their only apparent major disadvantage is that they are subject to actinic degradation. However, even the lower modulus version of aramid produces a stiff fabric with low failure strain.
Table 6.2:
Material
Polyester e.g. Terylene,
Dacron
Typical fibre properties of potential load bearing components of laminate materials Fibre
Fibre
Fibre
Strain to
Specific
Tensile Strength
Modulus
Failure
Gravity
GPa (Ib/in’)
GPa (Ib/in’)
ee Marr Woes 1.0
12
(1.2 x 105
(0.7 x105
Polyamide e.g.
Nylon
Aramid - high modulus e.g. Kevlar 49
e.g. Kevlar 29
;
3.9 x 109)
18.8 x 109)
(3.9 x 109)
(8.7 x 109)
Aramid - low modulus
10-15
WSWE*
Materials
\47
It appears that an aramid laminate (or coated material) is not able to disperse stress
concentrations which are inevitably introduced during the manufacture of a large structure such as an envelope. The unrelieved stress concentrations can then initiate catastrophic tear propagation. Only when this problem has been overcome or eliminated is there a prospect of stronger and lighter aramid envelopes. The fibres made from high performance polyethylene (e.g. Spectra and Dyneema) are very different from the material familiar as polythene. In polythene, the long polymer chains are curled up which allows the material to stretch several hundred
percent before failure. In the new fibres, the long chains have been straightened and lie along the fibre length, giving much greater strength with much lower failure strain. They are, however, too new to have built up any long term experience of their use although they have the potential to produce the lightest fabrics for a given strength, much less than half the weight of polyesters or polyamides. With similar strain to
failure and modulus as aramids, they may be expected to suffer from the same inability to relieve induced stress concentrations, which could initiate catastrophic tear propagation. Polyimides (Nomex) are known for their exceptional fire resistant properties (not a property normally specified for a helium filled envelope) but in other respects demonstrate similar properties to polyesters, but at a higher cost. Fabrics made from woven glass fibres are used extensively in the air supported building and tensile structures field, usually coated rather than laminated. They produce very stiff fabrics which require careful handling as they are prone to damage
when creased. It is not believed that they will produce a viable envelope fabric. Gas Retention
Component
The gas retention component can come as either a coating or a film. All the early materials used coatings as the gas barrier, as do materials now used for ballonets, although with the modern laminates, a film is incorporated as the gas barrier. The properties of materials that have been, or could be used as the gas retention component are given in Tables 6.3 and 6.4 for coatings and films respectively. Just as the other components can contribute to minimising helium loss, the gas retention component has more than a single function. In addition to low permeability,
it must possess shear stiffness to be able to enhance the overall stiffness of the laminate, and be able to be securely bonded to both the other components (being
sandwiched between them). Poor adherence to either could result in delamination. In practical terms, polyester again is the preferred material, with polyester film (Mylar - Du Pont) being the most common component. Polyester film has low permeability and is relatively strong and stiff.
148
Airship Technology Table 6.3a:
Typical properties of coatings potentially suitable for envelope materials Specific Gravity
:
0.
(Shore A)
Tensile Strength
30-100
(3,000)
Resistance
23
0. 92
Chlorosulphonated P.E (Hypalon)
1.12-1.28
50-95
(3,000)
Excellent
Polyester Elastomer
1.17-1.25
40D-72D
3,600-5,500
Excellent
Polyurethane
1.05-1.30
35-100
(4,000)
PVC
1.20-1.35
40-90
(1,500-
(Hytrel)
Excellent
3,500)
Table 6.3b:
Typical properties of coatings potentially suitable for envelope materials Heat Sealable
Natural Rubber
‘al
Weatherability Poor
Chlorosulphonated P.E (Hypalon) Polyester Elastomer (Hytrel)
Polyurethane
Good (with additives)
| Adhesion to Fabrics
Materials
Table 6.4a:
Material
Typical properties offilm materials for possible use in laminated materials Tensile
Ultimate
Strength
Elongation
7,000-16,000
30-60
Gas
| Permeability
Adhesion to
| Fabrics/Film
Polyurethane Polyvinyl Fluoride
PVDC Copolymer
Pair
(Saran) Poor
Low Density Polyethylene
Table 6.4b:
1,000-2,300
Typical properties offilm materials for possible use in laminated materials
Material
Flex
Dimensional
Fatigue
Stability
Resistance
Polyurethane Polyvinyl Fluoride
PVDC Copolymer | (Saran) Yes (some
grades only) Low Density Polyethylene
149
150
Airship Technology
Weathering Component
Some of the materials that have been considered as a gas retention component are also an effective weathering component, but one material stands out from the others, having been used for many years as a protective film. Polyvinylfluoride (PVF) film, commonly known by its trademark, Tedlar (Du Pont) has proven to be virtually inert at ambient temperatures to a wide range of acids, alkalis and solvents and has far
greater resistance to actinic degradation than most synthetic materials A comparison of fabrics with laminated PVF film against coatings used previously is shown below. Neoprene
e
SG 50% greater than polyurethane or PVF.
e¢
Modest weatherability gives 3 year life with frequent maintenance i.e. painting.
e
Must be joined with adhesive.
Polyurethane:
e
Very low helium permeability.
e
Very good handling properties and crease resistance.
e
Good weatherability gives 5 year life with modest maintenance.
e
Very easy to handle and join (by adhesives or heat bonding). Polyvinyl fluoride
e
Very low helium permeability.
e ~— Resists fungal growth.
e
Excellent weatherability gives 15 to 20 year life with no maintenance.
The very large increase in predicted envelope life shows the superiority of PVF
laminated materials, especially with the zero or minimum maintenance. This extended life coupled with the minimum maintenance greatly offsets the increased material and envelope manufacturing costs and a rigorous analysis would be expected to show a
much reduced life cycle cost for an envelope.
Materials
151
Bonding and Joining
The bonding agent that holds the laminate together is also the agent responsible for making joints and is thus a very important component of the laminate. A meltable bonding agent is specified both to aid in the manufacture of the material, and to improve the reliability of the joints. No methods exist to inspect joins, whether heat sealed or adhesively bonded, but the greater control in making heat sealed joints over adhesive bonded joints improves the reliability of the heat bonded joint. The choice of material with these capabilities is very limited. One material that has been successfully used is a polyester elastomer manufactured by Du Pont under the trade name ‘Hytrel’. This material has the properties required as a bonding agent, and also makes a contribution to the low permeability of the laminate. The nature of these bonding agents requires that aggressive solvents are used during the manufacture of the laminate. This has caused some problems in the past as certain of the solvents used have been banned by Health and Safety Regulatory bodies, necessitating the sourcing of usually more expensive replacements.
corr:
Environmental Protection Layer
txtrg
Helium Retention Layer
(=a
Load Bearing Component
Woven Butt Strap Bonding Tape Adhesive
Figure 6.2.
Section through typical laminate material joint.
A typical laminate joint is shown in Figure 6.2. The material is butt joined with a specially made tape as the butt strap, the joint being heated and cooled under pressure.
This strap carries all the load across the joint, while the outer strip of weather resistant The material is bonded there purely to seal the joint against the environment. and width tape used as the buttstrap is specially woven to have high strength across its
152
Airship Technology
is impregnated with the bonding agent. Similarly other special tapes in the form of ‘T’s or ‘X’s, also impregnated, are used for making special joints or attachments to the envelope. As with the polyurethane coated materials, the joints in the laminated materials are stronger than the materials themselves. In testing, the normal failure is found to be in the material, not in the joint. Potential of Laminate Materials
The laminated fabrics have the potential to produce very high strengths at low weight. For the US Navy YEZ 2A design which has a nominal envelope volume of 70,800 m° (2.5 million ft*), a single ply Dacron, laminated with Mylar and Tedlar using Hytrel as the bonding agent, was developed which gave a uniaxial strength of 5.25KN/50mm (600|bs/in) at a weight of 440 g/m? (13 oz per sq. yd). Derivatives of this fabric are under development which have the potential to achieve a uniaxial strength of 14kN/50mm (1600 Ib./in) at approximately 1 kg/m? (30 oz per sq. yd). It is believed that a non-rigid airship with an envelope of 1.1 million m® (40 million ft’) is feasible with a material of this strength (Munk). Ballonet Materials
The preceding sections have considered envelope materials, ie. the hull materials. The ballonet, however, is an important component of the envelope and the material from which it is fabricated requires an unusual combination of properties. The ballonet is continually inflated and deflated, moves against the inside of the envelope and itself, and during flight manoeuvres is subject to ‘ballonet slosh’, which can induce loads into the material, although since there is no pressure differential between the bailonet air and envelope gas, there is no pressure induced membrane stress in the ballonet material.
Ballonet materials therefore are required to be able to resist continual flexing and abrasion without increase in permeability, and to have moderate strength. Both coated
and laminated materials have and are still being used. The coated material is a lightweight nylon or polyester scrim coated on both sides with polyurethane; in some materials the coating is spread coated on one side and film transfer coated on the other. The combination of coating material, coating process and open weave scrim gives a very supple material with good flex properties that are maintained during extensive flex testing. The soft surface that goes with the suppleness would be expected to have greater friction when rubbing against itself or
the envelope, but in practice, must have sufficient abrasion resistance as abrasion problems have not been reported.
The laminate materials have an additional thin 0.01mm (0.25 mil) film of polyester bonded to the polyurethane coated, closely woven fabric. The combination of close weave and the film give a much stiffer fabric but with lower friction characteristics on the polyester laminated surface. Manufacture of the material itself is more difficult.
and joining is also not as easy as the polyurethane coated materials. While the flex
Materials
153
properties are not as good as the coated materials, as would be expected from the greater material stiffness, the laminated materials form perfectly serviceable ballonets with no reported service failures. The inference must be that the flex test used, while showing differences in material flex properties, does not sufficiently represent the flex conditions experienced by a ballonet in service. The materials are assumed to require only moderate strength since they are not subject to a continuous membrane load resulting from a pressure differential across the ballonet material. The ballonet, however, is subject to ‘slosh’ where the ballonet is observed to move violently under certain flight manoeuvres and significant loads in the ballonet material could be encountered. To my knowledge, no measurements of these loads have been made.
PROPERTIES OF ENVELOPE MATERIALS The design of an airship envelope is affected directly or indirectly by a large number of material properties, and very often compromises will need to be made over conflicting issues. In this section, the background to a number of the more important properties is considered. The preceding sections, however, have shown that the coated or laminated envelope materials are, in reality, complex manufactured structures. In conventional engineering structures, we can relate the behaviour of the manufactured structure, for instance a roofing truss, to the properties of the materials of which it is made, and its behaviour is critically dependent on its design. The behaviour of laminated envelope materials usually cannot be directly related back to the properties of their component parts, and we ascribe the term ‘property’ to describe the macroscopic behaviour of the complex manufactured structure. Not surprisingly, small changes in the manufacture or structure can have significant effects on these ‘properties’. In general, specific figures of the material properties discussed in the following section have not been given since they could be atypical and in the worst cases, misleading. It should also be remembered that in addition to the work openly reported, many ofthe values are unreported as they are considered to be commercially sensitive by both the manufacturers and users. Stress - Strain Properties
The uniaxial tensile strength, is probably the property most often quoted for an envelope material, and while a very important parameter, it gives limited information
about the anisotropic non-linear stress - strain properties of broadloom fabrics, whether coated or laminated. Further, modern stress modelling techniques capable of catering for non-linear materials require detailed data on stress - strain characteristics of the materials over the complete range, including bi-axial loading conditions.
154
Airship Technology
Biaxial Loading
The testing of woven fabrics under biaxial stress conditions has never been an easy test to carry out and biaxial testing is not a common standardised test. In recent years, however, the methods have progressed and several test machines, capable of producing a biaxial stress field by superimposing two orthogonal stress fields, have been
described
with the detailed
test procedures
required
(Reichardt,
Woo
and
Montgomery, 1953; Niwa, Kawabata and Kawai, 1971; Reinhardt, 1976). The test machines are complex instruments; one such that gave repeatable results was described thus, “...having a rigid frame 3m x 3m and ten hydraulic pistons on each side which are transversely movable and independently adjustable so that inaccuracies in the specimen can be equalised. The specimen has a square field of 50 cm x 50 cm... “(Losey 197.1), Measurements of biaxial stress - strain behaviour using this type of equipment have been carried out to validate theoretical predictions of behaviour based on simplified models of the fabric structure and properties of the fibres (Kawabata ef al., 1973) or to validate finite element models of cruciform test specimens under biaxial loading against practical measurements (Minami and Nakahara, 1981).
The generation of biaxial stress - strain data is a time consuming task since it has been normal practice to measure stress - strain characteristics in one direction while the orthogonal direction is maintained at constant load (or strain) To gather data over the full range of biaxial stress conditions requires a large number of repetitive tests. A cautionary note must be made: it has been my experience that in carrying out biaxial testing on nylon fabrics, the hysteresis effects are very significant and reproducible results could be obtained only if the test sample was allowed to recover for 24 hours after undergoing a single test cycle (Render and Bradley, 1986). If this is also the case for coated or laminated materials, the generation of full range biaxial test data could be lengthy and costly, although simplifying methods of predicting biaxial
stress-strain properties from limited data have been shown to be effective (Kageyama et al., 1988).
Shear Properties
Any envelope will be subject to significant shear loads, and any stress modelling technique will require shear strain data. As there is a need for reliable biaxial stress-strain data, similarly there is a need for reliable shear strain data. There are models of shear deformation of woven fabrics (Kawabata, 1989) and even of “fibrous assemblies” (Pan and Carnaby, 1989) such as non woven fabrics, based on the bending, rotation or slippage of crossed fibres or yarns. For uncoated cotton fabrics, a close correlation between measured and predicted shear deformation
properties has been found, the measurement method being based on an adaptation of a complex biaxial test machine (Kawabata, 1989).
The coating of a fabric, however, or bonding the fabric to a film, as in the laminate materials, drastically changes the shear behaviour, significantly increasing the shear
Materials
155
strength and stiffness. Measurements of shear properties have been made, usually based on inflated cylinders. In one study (Alley and Faison, 1972), a pressurised cylinder made of polyurethane coated nylon fabric could be independently, axially and torsionally loaded so that the circumferential, longitudinal and shear stresses could be independently varied, with deformations being recorded by photographing a grid drawn on the cylindrical surface. These measurements were claimed to be valid for increasing load only, since for cyclic loading the stress - strain characteristics were found to be explicitly dependent on the preceding load history. The measurements on coated fabric described above are not new and it would appear that little has been published recently. In general, however, if particular shear data are required by the designer or stress analyst, methods do exist whereby this may be obtained. Permeability
Envelope materials do not form membranes impervious to gases. Not only will helium diffuse from the envelope, but, in theory at least, atmospheric gases and water vapour will diffuse into the envelope since diffusion of a gas through a membrane is dependent on the difference in partial pressures of the gas on either side of the membrane. Helium is a monatomic gas and has the smallest molecular diameter of all gases with a consequence that, under identical conditions, it will diffuse through a material more rapidly than any other gas. In contrast, the atmospheric gases, mainly oxygen and nitrogen, have much greater molecular diameters, and any diffusion into an envelope is assumed to be very small - an assumption that operating experience validates. As a comparison, the permeability of nitrogen through PVF is
approximately 0.05% ofthat of helium so that helium loss would outstrip any dilution effect by a large margin. The loss of helium from an envelope is a major consideration from an operational viewpoint. It cannot be prevented since helium will diffuse through any envelope material - even as it must have done through the aluminium alloy skin of the metal clad US Navy airship ZMC-2.
Helium loss has two effects: a)
it leads to loss of lift and hence a reduction in operational capability of the airship; this is rarely a safety issue since the loss is normally at a known rate and can be accommodated.
b)
it adds to the operational costs since to maintain the operational capability, the helium must be replaced.
While as envelope size increases, the square/cube laws ensure that helium permeability has a diminishing effect on operations, (the helium loss reduces as a fraction of the total helium fill), in financial terms, the larger the envelope the greater
is the volume of helium lost for a given permeability. The need to minimise helium
loss has always been present, but the approach has changed from increasing the
156
Airship Technology
weight of rubber or coating the envelope with paraffin to choosing materials with the least permeability to helium. Because the measurement of helium permeability, using standard methods, was found to be variable and the measurements did not correlate with the achieved daily lift loss TCOM instigated a study into the measurement of helium permeability in
their laminated envelope materials (Ashford et al., 1983). The permeability of these laminated materials (or more strictly permeation, since the term ‘permeability’ is applied to single materials while ‘permeation’ is applied to composite materials) was measured by independent test laboratories with large variations in the results obtained from nominally the same test apparatus and methods. By careful attention to experimental detail and methodology, they succeeded in developing a reliable and repeatable test method. From the permeability of the laminate components, as supplied by the material manufacturers, it was possible to calculate a theoretical value for the permeation of the trilaminate i.e. the Tedlar and Mylar laminated to Dacron with Hytrel as the bonding agent. This value was found to be 1.14 I/m°DA (litres/[square metre|[day][atmosphere]). and compared well with the measured value of 0.99 V/m?DA. Of more practical interest, they calculated the theoretical lift loss of a 411,000 cu ft aerostat, based on the measured permeation value, to be 12 Ib. per day against the lift losses of 14 to 16 Ib. per day measured from several aerostats. This would lead to the conclusion that helium losses, often believed to be the result of leakage through hardware or pinholes, is, in the new generation of envelope materials, almost entirely due to permeation of helium through the envelope material. Further reductions in permeation can be achieved only through the selection of an improved gas barrier film and since it is possible to predict the permeation from the component materials permeability, the search for a better gas barrier film can be concentrated on those with known low permeability. Tear Resistance
The
new
generation
of envelope
materials
are
tougher
than
those
used
previously, but this is difficult to quantify as the methods of measuring tear resistance that lead to a quantitative prediction of envelope tear behaviour are not fully developed and to my knowledge none of the data on laminate materials that has been published relates to catastrophic tear propagation - amongst the most critical features
of an envelope material. The background to tear measurement gives an idea of the difficulty in obtaining adequate data. Measurement of Tear Strength
Methods of measuring tear strength and resistance have been developed over
many years. Harrison (1960) published a comprehensive literature survey into the tearing strength of textile materials, containing over 200 references spanning the years 1910-1960.
Materials
157
Quantitative measurements of the tear strength of textile materials go back to the early decades of this century, not surprisingly when it is considered how important textile materials were in the manufacture of aircraft. Tearing was considered a major failure mode in textile materials both from direct tearing resulting from snagging and for propagation of a tear from damaged or holed areas e.g. a bullet hole in the fabric covering of a wing. In addition to aircraft fabrics, tearing strength was cited as important in fabrics used for parachutes, tentage, wagon covers and even industrial fabrics used for bags and sacking. A large effort was therefore put into methods of measuring tear strength and many of the methods still used to this day were developed during World War I. The wounded (slit) tensile test, single rip and tongue tear tests, trapezoidal test and wounded (slit) burst test were developed and several are used to this day. All of these tests, while effective in giving comparative tear strengths of textiles, do not give results that can predict how a tear in a textile will behave in a given situation, although some attempts were made to define conditions under which tear propagation would occur. The tests did allow qualitative effects to be outlined, all the analyses being based on the assumptions that:
e
in tearing, all threads are broken singly and tear strength is roughly proportional to the single thread strength of the threads broken;
e
although broken singly, several threads adjacent to the next thread about to break, are put under tension by the tearing force;
e
the number of threads tensioned this way is determined by the thread spacing.
Based on these assumptions, qualitative statements about the effect of different fabric features were made, generally based on how these features affected the load carried by the adjacent threads ahead of a tear. These are summarised below: e
Yarn smoothness (or roughness) through the friction between crossed threads, and extensibility affect the loads that can be carried by threads adjacent to the next one due to break (i.e. load concentration ahead of the tear).
e
Yarn twist has a secondary effect through changing yarn smoothness and thread spacing.
e
Increased thread strength through use of a coarser yarn affects tear strength, but
while even a small increase in coarseness could substantially increase the rip strength, a considerable increase in coarseness is necessary to increase the wounded tensile strength.
e
The effect of thread spacing, being inconsistent features, is secondary.
and depending
on other
158
©
Airship Technology
Crimp has two effects: a higher crimp gives a more extensible fabric, and thus a wider distribution of load ahead of a tear and higher tear strength; but a
reduction in ease of slippage because of the greater bend in higher crimped
threads leads to a lower tear strength. In the wounded burst test, ensuring that warp and weft (fill) extensibilities are equal, by controlling the crimp balance, burst pressures are improved. e
Weave affects the sleaziness of the fabric, the ease of thread slippage and the number of threads breaking together, and the superiority of matt-weaves over plain-weaves was shown.
e
Finishing processes such as scouring, bleaching or dyeing affected the tear strength either by altering the thread strength or altering the thread smoothness and thus effects were secondary. Processes, however, such as mercerisation and
the use of water repellent waxes affected ‘yarn lubricity’, thus promoting thread slippage and distribution of loads. e
Coatings, all lumped together as having a similar effect, were shown to reduce the tearing strength by an amount dependent on the restriction of thread movement.
Harrison's survey is a very comprehensive summary of the state of knowledge current in 1960. The bulk of work surveyed concentrated on the rip and trapezoidal tear tests using uncoated fabrics. In these tests, the load applied directly to the tear to cause it to propagate was measured, while the fabric away from the tear is unstressed. If the load was removed the tear ceased to propagate. Tearing mechanisms proposed in the analysis of tear propagation assumed that threads broke individually, and that threads close to the tear were loaded. Quantitative analysis on this type of tearing closely described the behaviour in terms of basic fabric and yarn properties.
In both wounded tensile and wounded burst tests, however, a different type of behaviour was observed. The test samples were evenly stressed under increasing load, and at some load the wound began slowly to propagate. At this stage, if the load was removed, propagation ceased. At some slightly higher load, however, tear propagation became catastrophic, the tear propagating at some hundreds of feet per second. While it was believed that the same principles applied to this behaviour as to conventional tearing, no theoretical analysis of rapid tear propagation had been carried out.
It is possible that this mode of behaviour was not perceived to be of major importance; according to Abbott and Skelton (1972), catastrophic propagation is more likely to be seen in coated fabrics, a topic scarcely touched by Harrison's survey. Accordingly, none of the effort carried out before 1960 can help in determining the susceptibility of large stressed fabric membranes to the catastrophic extension of an
existing tear. While new approaches to failure have been developed that can be applied to tear propagation in textiles (and textile based materials), the general concept of tear behaviour is still heavily influenced by the approach described by Harrison.
Materials
159
Fracture mechanics approach
The recognition that all materials and engineering structures contain flaws or defects, that under certain conditions can propagate catastrophically, has lead to the development of fracture mechanics. Initially developed to explain the behaviour of brittle, high strength materials, its application has been extended to the behaviour of ductile materials. The fundamental of fracture mechanics is the definition of a function known as the stress intensity factor. This function describes the stress field concentration around the tip ofa crack (or end of a tear) in terms of the overall stress and crack length. The function, when conditions are linear elastic, takes the form: of = K/(2a)0.5
(6.1)
where
Of K a
= = =
applied stress stress intensity factor crack semi-length
Below a certain value of the stress intensity factor, the crack will not propagate or propagate only slowly. Above this value of the factor, crack propagation is
catastrophic. This critical value, Ke, has been shown to be a material constant and can be considered to be a measure of material toughness. The fracture mechanics approach recognises different forms of crack opening: Type I is tensile opening of the crack in response to tensile load across the crack in the plane of the material; Type II is a shear opening of the crack in response to a shear stress along the crack, in the plane of the material; Type III is a shear opening of the crack in response to a shear stress across the crack, normal to the plane of the material. Until the advent of fracture mechanics approach, most of the investigations into the tear resistance of textiles have been carried out using some form of tear test where the
tear opening is of Type III and national specifications for these types of test are still current. In contrast, the type of failure seen in stressed textile structures is Type I tear
opening where at some point the tear extension becomes uncontrollable. Considering this difference in mode, it is not surprising that conventional tear testing has given no information on the response of an envelope to a tear. Conventional tear testing may have its uses as a quality acceptance or monitoring test, but even this is debatable since tear strength correlates well with single thread strength and any changes would be reflected in the textile tensile strength.
160
Airship Technology
Application of Fracture Mechanics Approach to Textiles During the 1970s and 1980s the principles of fracture mechanics have been applied to coated textile materials. The impetus for this work has been primarily driven by the use of high strength coated textile materials in two areas: air supported structures (air houses) and tensile structures for large span roofing. In both these types
of structure, uncontrolled
propagation of a tear initiated at a flaw could have
disastrous results and since the materials
used are susceptible
to foreign object
damage e.g. objects blown by high winds, it is probable that the stressed materials will sustain damage at some time (Ansell ef al., 1984). As in an airship envelope, the initiation of the tear is far less important than its propagation characteristics. A growing body of literature exists which addresses the problem of tear propagation. Findings are summarised here but the reader is advised to consult the original papers for more details Earlier attempts were made to measure tear propagation (Kawabata, 1989) or critical tear lengths in pressurised fabric cylinders (Topping, 1973) and provide theoretical assessments. These studies recognised shear stiffness of the material as an important factor, since it is through shear that the stress concentration at the tear tip is diffused away. There have been attempts to derive a theory for tear propagation in coated fabrics based on the propagation of a tear in a uniaxially or biaxially loaded test specimens (Minami, 1978; Williams and Gaafar, 1984), and to correlate test results with theory (Minami, 1978). Minami (1978) developed a theory based on Hedgepeth's model of stress concentration factors due to cracks in fibre reinforced composite materials (Hedgepeth, 1961) using a form of Griffith energy balance as a criterion of failure. From this unlikely beginning, the function he developed as a measure of toughness is essentially constant for varying crack lengths, a necessary condition for a measure of material toughness. One aspect of these theories, however, is that they all assume that the stress - strain characteristics of the coated fabrics are linear elastic, while as has been pointed out, the materials are anisotropic and non-linear. I am not yet aware of any developed theories that have taken the non-linearity into account, although it has been pointed out (Racah, 1984) that in the general equation given above, the exponent of a is equal to 0.5 only under linear-elastic conditions. Both theoretical analyses and practical studies indicate the importance of the stress concentration at the tear tip and the way that the stress concentration is distributed ahead to the tear (c.f. the plastic zone ahead of a crack in a ductile metal). The stress is carried by shear between adjacent threads and size and distribution of the stressed zone is affected by the shear stiffness or modulus of the material. The greater stiffness
of the laminate materials appear to enable the stress concentration to be diffused over a greater distance with a consequently improved tear resistance.
Materials
161
Uniaxial and Biaxial Tear Testing While most tests have been carried out under uniaxial tension, a growing number have been carried out under biaxial tension. It has been shown that the load at which tear propagation starts is less under uniaxial tension than under biaxial because "since the lateral deformation which is caused in the vicinity of a defect under uniaxial stress is suppressed in the case of biaxial stress, the magnitude of concentrated stress in the vicinity of the defect becomes smaller in the case of the biaxial stress condition" (Minami and Motobayashi, 1981), a finding that has been theoretically predicted (Williams and Gaafar, 1984). This finding does suggest that tear propagation data obtained from uniaxial testing is
sufficient since the results found are pessimistic relative to the more complex and more costly biaxial tear testing. The standard uniaxial slit tensile tests, however, may give unrepresentative results if the sheared area, which carries the load away from the tear tip, is interfered with by the test machine grips or the edges of the test sample. Small test samples as found in the standard tear strength tests are probably of insufficient size and it has been proposed (Racah, 1984) that for realistic tests, the test sample length should be a minimum of 20 times, and the width 5 times, the initial tear length. The test sample. therefore, is not necessarily a constant size, and large samples, up to | metre square, may be required. The problem of tear propagation in coated and laminated textiles has not yet been solved. Sufficient progress, however, has been made to identify important factors and
give confidence that further effort in this topic will provide a practical measure of material toughness. Flexural Properties
It has been stated previously that ballonet materials must be able to resist continual flexing and abrasion without increase in permeability. While easily stated in general terms, in practice, it is more difficult to define what is required as a test of these properties and to lay down acceptance criteria - what level of a parameter 1s or is not acceptable. In addition. it is often the permeability that is measured after a sample of material undergoes a fixed flexural regime intended to simulate the flexural conditions expected in practice. Consequently a variety of tests have been developed specifically for both envelope industry and ballonet materials that are unused by any other industry, while other specific tests have been borrowed.
A relatively early flexural test is the Rotoflex test specified in US military airship at material specifications. In this test, a circular specimen is gripped across a diameter for other a fixed load, and one grip rotated + 270° for 200 cycles This is then repeated this After . directions thread the on diameters, the new orientation depending the nor limit, given a treatment, the permeability must not have increased beyond
162
Airship Technology
strength dropped below a percentage of the untreated strength. This procedure can be seen as an attempt to simulate the flexing undergone by the materials in practice, but
is more
of a conditioning process carried out before measuring
the materials’
permeability and strength. A conditioning procedure aimed primarily at the handling properties of aerostat materials is called Twist-Flex. This simulates a very severe handling regime. Two interlinked cylinders of material, approximately 0.3m diameter by 1.2m long, with a common axis are alternately inflated and deflated. The inflating cylinder twists the deflating cylinder through half a revolution while crushing it from 1.2m to 0.2m in length. The process is then reversed, the air being exhausted from the inflated cylinder and pumped to the deflated cylinder. The normal handling of an aerostat including manufacture and deployment is said to be represented by 100 twist-flex cycles. A variety of tests, several coming from the leather or shoe industries, have been
adapted and used by the airship manufacturer or their material suppliers, with each one having both advantages and disadvantages, and some have been developed into national specifications (ASTM D2097-84, BS 3424:Part 9:1980). The specific test employed is probably unimportant if the test is being used purely comparatively, or the user has carried out sufficient testing to correlate the test results with resistance to damage by flexing that his application requires.
IMPROVED LAMINATES
A recent potential addition to the range of materials from which envelopes could be fabricated has emerged - laminates which have been developed as sailcloth for high technology racing yachts. While the basic construction of these laminates is similar to those described earlier, the major difference is that the load bearing component is a scrim. Normally laminated between two polyester films, the scrim is not woven and is made from yarns with neither twist nor crimp. The yarns are laid in the warp and fill directions,
and in any required bias direction. By a combination of different weights and different types of yarn, the material can thus be manufactured to give different desired strengths and stiffness’ in different directions. The laminates are claimed to be light, stiff and strong, with excellent tear resistance. While extensive test programmes must have been carried out during the development of these materials, it appears that only very limited information on the laminate properties has been published. Nevertheless, the laminates do appear to have
the potential of being developed into ideal envelope materials.
SUMMARY
This chapter has attempted to give some insight into the properties of textile based materials from which an airship envelope can be made.
Materials
163
While major developments in textile engineering and science have transformed the
capabilities of these materials, some important properties and characteristics are still poorly understood and their impact on envelope design not fully recognised. I believe a common mistake is to consider a woven textile material, whether coated or laminated, as an isotropic thin sheet. It should be remembered that textiles are anisotropic, have non-linear stress-strain properties, and their orthogonal stress-strain properties interact through crimp interchange. Having said that, they are engineering materials that behave according to fixed rules, even if the rules may occasionally be somewhat obscure or not fully understood.
REFERENCES Abbott N.J. and Skelton J. Crack Propagation in Woven Fabrics. J. Coated Fibrous Materials. 1, 234, 1972. Alley V. L. and Faison R. W. Experimental Investigation of Strains in Fabric under Biaxial and Shear Forces. J. Aircraft. 9, 55, 1972. Ansell M., Barnes M. and Williams C. Structural Property Tests for Coated Fabrics. Institute of Structural Engineers/University of Bath Conference, The Design of Air Supported Structures, Bristol, 1984. Ashford R.L., Bata B.T. and Walsh E.D. Measurement of Helium Gas Transmission
Through
Aerostat
Material.
AJAA_
Lighter-Than-Air
Conference,
Anaheim,
California, 1983.
ASTM D2097 -84. Standard Method of Flex Testing of Finish on Upholstery Leather. BS 3424 : Part 9 : 1980. Testing Coated Fabrics. Part 9. Methods 11A, 11B, 11C and 11D. Methods for determination of resistance to damage by flexing. Harrison P. W. The Tearing Strength of Fabrics - A Review of the Literature. J. Textile Institute. p61 Mar 1960. Hedgepeth J. M. Stress Concentration in Filamentary Structures. NASA Tech. Note D-882, May 1961. Houmard J.E. Maximum Size of a Nonrigid Airship. AAA Aircraft Systems, Design and Technology Meeting, Dayton, Ohio, 1986. Kageyama M., Kawabata S. and Niwa M. The Validity of a "Linearizing Method" for Predicting the Biaxial-Extension Properties of Fabrics. J. Textile Institute 1988,
No. 4. Kawabata S., Niwa M. and Kawai H. The Finite Deformation Theory of Plain-Weave Fabrics. J Textile Institute 64, 21, 1973. Kawabata S. Non-Linear Mechanics of Woven and Knitted Materials. Textile Structural Composites. (Composite Materials Series Vol 3) Piper R. B. (Editor) Elsevier 1989. Losch M. Bestimmung der Mechanischen Konstantern fur Einen Zweidimensionalen, Nichtlinearen, Anisotropen, Elastichen Stoff am Beispiel Beschichteter Gewebe. Diss. Univ. Stuttgart 1971.
164
Airship Technology
MIL-C-21189(Aer). Cloth, Laminated, ZPG2 and ZPG2W Type Airship Envelope. 15 July 1959 Minami H. Strength of Coated Fabrics with Crack. J. Coated Fabrics, 7, 269, 1978. Minami H. and Motobayashi S. Experimental Research on Uniaxial and Biaxial Tensile Strength of Coated Fabrics with Variously Shaped Defects. J. Coated Fabrics, 11, 24, 1981.
Minami H. and Nakahara Y. Experimental Verification of the Method of Deformation Analysis of Coated Plain-Weave Fabrics by Biaxial Tensile Tester. J. Coated Fabrics. 11, 11, 1981.
Munk R. Private communication Niwa M., Kawabata S. and Kawai H. Analysis of the Anisotropic Tensile Properties of Plain-Weave Fabrics. J. Textile Machinery Soc. Japan. 17, 47, 1971. Pan N. and Carnaby G. A. Theory of Shear Deformation of Fibrous Assemblies. Textile Research J. 59, 285, 1989.
Racah E. Crack Propagation Testing of Coated Fabrics Used in Surface Stressed Flexible Structures. J. Industrial Fabrics, 3, 4, 1984. Reichardt C.H.,Woo H.K. and Montgomery D.J. A Two-Dimensional Load-Extension Tester for Woven Fabrics. Textile Research J. 24, 424, 1953.
Reinhardt H.W. On the Biaxial Testing and Strength of Coated Fabrics. Experimental Mechanics. p71, Feb. 1976.
Render A. B. and Bradley P. D. The Development of a Parachute Strain Measurement Technique. AIAA.....Conference, Albequerque, 1986. Topping A. D. The Critical Slit Length of Pressurised Coated Fabric Cylinders. J. Coated Fabrics. 3, 96, 1973.
Williams C. and Gaafar I. Tear Propagation in Coated Fabrics. Institute of Structural Engineers/University of Bath Conference, The Design of Air Supported Structures, Bristol, 1984.
7 pe
Structures C. Luffman
INTRODUCTION Airship structures traditionally fall into four groups: rigid, non-rigid. semi-rigid and hybrid. Rigid Airships normally have an overall skeletal frame. to hold their form (carrying overall bending), which is covered in light dope-tautened material (to produce an aerodynamically smooth outer profile), into which are fitted separate gas containment cells. All other systems and structural features are carried by the skeletal frame. An exception to this class of airship are those which utilise an outer monocoque shell, stabilised by internal gas pressure (similar to the ZMC-2 metal-clad. built in 1929). This latter type may also be referred to as a pressure airship. Non-rigids are also pressure airships, since they maintain their form by pressure
stabilisation of the gas containment membrane (the envelope) - similar to aerostats. The membrane material is normally a flexible gas-tight and weather-proofed fabric, enabling it to be folded when the envelope is deflated. They are sometimes referred to as blimps. They do not have a skeletal frame, since the envelope under pressure acquires a residual membrane tension, becoming stiff enough to resist overall bending effects and to carry all other structural features without reducing the membrane tension to zero. Semi-rigid airships combine features from both types, essentially providing a rigid full-length lattice beam (normally a keel structure) and a pressurised envelope, which act together to carry the effects of overall bending and maintain shape. In a keel structure of sufficient proportions, the lattice beam can replace the need for a separate car, providing the housing for the airship’s systems and the means to mount the propulsion and landing gear units. It stiffens the lower envelope, which suffers owing to low static head (reducing the stabilising gas pressure longitudinal tension); hence enabling hogging bending loads to be carried and spreading the weight ofthe installed systems over the envelope’s entire length (Pagon, 1927). These three types of airship are normally of conventional cigar shaped form, designed to minimise drag, enabling them to be directed through the air (1.e., dirigible) via aerodynamic surfaces and propulsion units. Hybrids are nonconventional types of airship, of special form, not necessarily relying on conventional methods for their means of lift, aerodynamic control, propulsive system or structural arrangement. Although general principles may be applicable to hybrid airships these are not specifically covered by this chapter.
166
Airship Technology
HISTORICAL
With advances in material sciences, joining techniques, analytical facilities, design
aids,
atmospheric
awareness
and
aeronautics
in
general,
there
is the
opportunity for efficient modern airships to be introduced at competitive prices. A number of lessons should be noted from the past, namely: Atmospheric disturbances must be respected and understood. Some airships have been lost owing to severe weather or turbulence (Althoff, 1990), including those at mooring masts or whilst passing through their hangar doorways. Tail surface loads must not be underestimated. The R38 (Kinsey,
1988), may
have been lost owing to structural failure for this reason. Airships need to be fully tested before entering service. This must not be rushed or side-stepped by those eager to impress or earn revenue before the new type has been properly endorsed by the certifying authorities; this may have been a contributing factor to the R101 disaster (Chamberlain, 1984).
Rigids tend to be fragile - unforgiving; while non-rigids are more flexible and resilient. For example, if the tail surfaces of a rigid were to strike the ground inadvertently, serious structural damage might follow; whereas a non-rigid’s envelope might deflect temporarily without failure - popping back into shape when the load is removed.
GENERAL PRINCIPLES AND CONSIDERATIONS This section introduces the reader to airship principles and aspects, which must be given structural consideration. Equilibrium Control
the
An airship’s structure must provide facilities essential to weight control over effects of buoyancy (gas lift). Whenever appreciable differences occur
compensating ballast must be added or dumped to enable the airship to maintain equilibrium. This may occur for several reasons, for example: Rain wetting or evaporating from the envelope.
Gas superheat changes from varying sunlight or other weather changes. Fuel usage.
Embarkation or disembarking.
Structures
167
As the overall capacity for compensating ballast is large to cover these aspects (about 15% of the gas lift), and its position affects balance; it has a major influence on the design of the airframe. Adequate structural provision must be made for the ballast’s support, containment, access, fill/dump, quantity measuring and anti-icing arrangements. The requirement for equilibrium control also affects the way servicing arrangements are executed. Special procedures need to be developed to transfer compensating weight between the airship and the ground as an item (such as an engine) is installed or removed, to enable such service arrangements to be performed. These details need
to be sorted out early in the design, since their influence affects the layout of the structure and provision of hard points for lifting.
In addition, for emergency escape purposes, the facilities and procedures must ensure the airship will remain on the ground following a forced landing. This probably requires a controlled but rapid means to dump the lifting gas. As buoyancy is lost, however, support of the airship’s weight transfers from its hull to the ground. The airframe must be designed to accommodate the redistribution of load in a way that ensures it settles to a steady position, without tipping over, blocking or affecting the escape positions, and without collapsing on top of personnel. As weight is transferred, the ability of the landing gear to withstand the resultant loads may be questionable. Since an airship can be fitted with only one support unit, pendulum stability normally keeping the airship upright, release of the lifting gas with such a configuration would also cause the gondola to tip over. The landing gear should, therefore, be designed to collapse gracefully, without puncturing the airframe, and/or be retracted as necessary. Motions at the Mast
An airship’s structure must be designed to accommodate the ceaseless motion arising from atmospheric disturbances when moored. Clearly this significantly affects the design of landing gear, mooring structure, handling features and their carrythrough structures. Loads and fatigue criteria must be established to accommodate the effects, the airship’s movements probably being uncontrolled. Limitations and procedures must also be established to ensure the airframe is not subjected to the effects of abnormal convective activity. An auto-stabilisation system, which provides auto-pilot control of the airship at the mast, may usefully limit excessive loads from being developed during gales. An important design case, which the airframe and installed systems must be designed to withstand, is the possibility of a kiting motion where the tail rises up to such an extent that the airship reaches an attitude with its longitudinal axis near vertical - appearing to perform a nose stand. Attitudes up to 70 degrees from the
horizontal plane have been recorded, as Figure 7.1 demonstrates. This may occasionally occur, whilst tethered, due to air turbulence - aggravated by poor location of the mooring site relative to surrounding buildings or regional of the prominences. The near vertical attitude is probably the critical case for design
168
Airship Technology
gondola’s longitudinal restraint method, since the loads resulting can surpass normal flight conditions. Flight attitudes may be limited to + 45 degrees. In addition, since the gas pressure varies with height, its container must be designed to withstand the resultant build-up due to pressure head.
Figure 7.1.
Airship Industries SKS 600 series airship at a mooring mast.
Structures
169
Unless hangared the airship must be designed to align itself freely with the wind, circulating around its mast (similar to a weather-cock) or rising from the ground in a pitching motion as prevailing conditions dictate. Otherwise the loads of restraint would be impossible for the airframe to bear. If suitable hangar facilities are not available (a rarity for airships at the moment) or weather conditions prevent the airship from being taken in, service arrangements may have to be conducted at the mooring site. As stationary ground equipment or vehicles are a hazard to the moving airship under such conditions, the airship either must be designed to pass safely over necessary obstacles or be restrained (bearing in mind that restraint is only practical during very calm conditions - wind strength less than 5 knots). Ground related tasks must, therefore, be executed from the airship instead of from the ground. For example, boarding steps must be attached to the airship without touching the ground. As a result, structural access and attachment positions for service equipment needs to be provided on airships that would not be required for heavier-than-air craft. Equipment temporarily installed on an airship may be needed for: boarding, loading, reprovisioning, fuelling, engine changing, and other maintenance actions. Accordingly, the equipment should be designed in concert with the airframe. to ensure the airship’s structure is provided with suitable attachments and can accommodate the loading actions. Boarding facilities must ensure safe personnel transfer, particularly as the airship may move without warning or rise up from the ground owing to atmospheric disturbance. The airframe also should be provided with convenient earth attachment points for use while the airship is held in proximity with the ground, to prevent electrostatic effects - essential whilst fuelling. Size Effects
The sheer size of an airship’s structure has a major influence upon its design. Lengths of members are limited by the capacity of tooling, transport and handling facilities, making additional joints necessary. Major sub-assemblies may need special transport joints to be engineered, so that they can fit into standard containers. Special handling positions also need to be provided to enable lifting equipment to raise them to their installed position. The overall arrangement of the structure is also affected by the hangar facilities
available for its construction and maintenance. Compromises affecting the length to
diameter ratio (for a particular lifting gas capacity), tail surface arrangement. propeller installation, landing gear length and gondola depth/support method. may need to be made in order to clear it through the hangar’s doors. This must also be taken into account when considering the prototype airship’s potential development.
Structural Fragility
While airships are usually larger than other types of aircraft, the airframes the four mass per enclosed volume or surface area is much less. For example, one of
170
Airship Technology
tail surfaces on a 2.5 million cubic foot airship (70 800 m3) - which would have an area of about 1300 sq ft (121 m2) and weigh approximately 1750 lb (794 kg), ie: 1.35
Ib/ft2 (6.57 kg/m) - may be compared with:
Glider
1.0 Ib/ft2 (4.88 kg/m2)
Light aircraft
2.4 Ib/ft2 (11.72 kg/m2)
Boeing 747
20.0 lb/ft? (97.65 kg/m?)
The delicate nature of an airship’s structure is therefore an issue which must be given careful consideration, especially for handling, shipping and other working practices. Minimum gauge for handling may often dictate the nature of the structure. Problems of panel or member stability also must be considered. To overcome such problems the design might adopt the use of braced frame structures with materials of high strength/weight ratio and high bulk content, skinned with light fabric as necessary. Where access is necessary, honeycomb stiffened panels can be used to provide a durable surface. Damage-tolerant structural design principles should only be applied as far as practical, recognising that for such vast structures it is not possible to make them damage resistant without increasing the weight by an inordinate amount. Fail-safe principles should be applied wherever possible. Vital parts, the failure of which may lead to catastrophic loss of the airship, should be protected against inadvertent damage, be inspectable and have stringent quality control procedures applied. Load Factors/Aerodynamic Pressure
When in flight an airship is a gentle giant; other aircraft, by comparison, move
speedily. As a result load factors and aerodynamic pressure distributions on the airframe are much lower. Maximum in-flight incremental acceleration load factors, owing to worst gusts or manoeuvring, may only be: 0.75g vertical, 0.5g side and 0.5g longitudinal. Even large passenger aeroplanes, such as Boeing 747s, designed to fly sedately are subject to significantly higher load factors. This difference must be
carefully understood, since it leads to much lighter accommodate handling rather than the loads of flight).
scantlings
(designed
to
Failure Effects
When an airship has a failure in flight - for example, total loss of power or the flying control system - it can revert to being a free balloon. Other aircraft probably would crash in similar circumstances. As a result, philosophy in design is often different. Airships, for example, can have their engines repaired in flight. This leads to additional access provision (which must be accommodated by the structure) and the
relaxation of some standards. Minor structural damage, such as holes in the gas container, might also have temporary repairs applied in flight.
Structures
171
Endurance Effects
Airships have the potential for much greater endurance than other aircraft. They can be designed to remain aloft for a week at a time without refuelling. With in-flight refuelling this can be extended to months before returning to base. This, of course, affects maintenance or inspection periods and consequently reliability goals for the design of the airframe, which may need to be of a higher standard on critical parts compared with heavier-than-air craft. If the airship is being designed to have such long endurance then facilities to enable inspection of the airframe in flight, along with access and essential safety line attachment positions, should be provided. Gondola Shear Force and Bending Considerations An airship’s gondola can have more or less continuous support along its length. An aeroplane’s fuselage, on the other hand, is cantilevered fore and aft of the wing. The different shear force and bending moment diagrams lead to structures that are fundamentally different. An airship’s gondola is lighter and can be much more uniform, with fewer stiffeners, since shear flows are smaller. A layout similar to a boat’s hull is not unreasonable, although every effort must be made to minimise weight. The nature of the gondola’s structure may depend on the altitude the airship is designed to operate at (leading to the need for cabin pressurisation) and whether it is designed to settle on water. Aerodynamic Drag
Owing to the relatively slow flight speed of an airship, minor protuberances from the skin line have much less effect on performance. Raised head fasteners which are structurally better than countersunk head fasteners - external skin straps and other such skin line disturbances can be considered for use when they would not be accepted on an aeroplane. This does not mean to say that external stringers or a corrugated surface would be acceptable, since skin friction drag would rise appreciably with the increased wetted surface area. Handling lines, tail fin bracing wires and external gondola support cables also create a significant performance loss (owing to their great lengths being positioned across the airflow), so should be retracted, aerodynamically shaped, or faired as necessary. Emergency Escape
Special emergency escape facilities may need to be devised because airships
can become airborne again (owing to buoyancy or bad weather conditions) and (depending on the nature of the incident, type of airship and whether the lifting gas and escapes or is released) their envelopes or hull may collapse over the passenger
172
Airship Technology
crew compartments. conditions.
The facilities must enable rapid yet safe escape
under such
For these reasons conventional heavier-than-air craft escape slides (which rely on the aircraft staying put following the incident and without debris obstructing their operation or the passengers evacuation) may not be suitable for airship use. Special lanyards, clipped to a safety harness, that pay out as the escapees jump from the exit positions (slowing their descent rate and being released when safely on the ground) have been considered for some airship proposals as an alternative. Depending on the airship’s role, however, the facilities provided must be suitable for the type, skill and number of people evacuating the occupied areas. It should also be understood that, in such circumstances, if gusting or windy conditions prevail, the airship may be dragged along, turning broadside to wind and perhaps tipping over, until sufficient lifting gas has been lost for the gondola’s weight to take effect. Emergency grappling lines may be required to hold the airship into wind as the lifting gas escapes. Sufficient escape positions that will not be rendered unusable, and the means for the escapees to make their way quickly to the exits (and operate them) whilst the gondola is on its side, need to be provided. For ditching on water, a sea anchor should be designed, which deploys from the bow to hold the airship into wind. The gondola’s structure must be designed to withstand the hydrodynamic and hydrostatic pressures experienced as it enters and sinks into the water. The gondola also must be designed with sufficient watertight buoyancy chambers to prevent it from sinking - should the lifting gas be lost - and hold it at an acceptable attitude long enough for escape purposes. Consideration should also be given to the problems that the rescue services would face under emergency conditions. Provision for means to keep the airship on the ground, make safe access into the occupied compartments and perform normal emergency functions - such as extinguishing fires - must be covered in the airframe’s design. A particular feature necessary is provision to release the lifting gas from an accessible position by an operator on the ground. Such provision must be clearly marked and protected from inadvertent use without restricting intended operation. The method must enable rapid release of the lifting gas under controlled conditions. If possible the method should also prevent the envelope from flopping over emergency exits, which otherwise could hamper further rescue actions. Clearly, implements to clear such debris rapidly are necessary and should be readily available for the rescue services’ and airship crew’s use. Weight, Mass Distribution and Balance
Weight limitations and overall balance are difficult to achieve, since small tolerance differences over such big structures have a significant effect. For example, if actual fin weights do not match their design estimates overall balance will be
affected owing to the fins’ extreme
aft position. It is also difficult to predict
accurately the centre of buoyancy’s position. This is because the gas container’s
shape (which is indeterminate, but idealised as a regular shape or body of revolution)
Structures
173
varies according to applied pressures and other loads, its materials of construction and inertia or gravitational effects. Inevitably, the difficulty of predicting these values accurately may result in a large balancing weight (penalising performance expectations) needing to be added at the nose or tail to bring the airship’s centre of gravity within acceptable limits relative to the centre of buoyancy. Consequently the nose and tail structures should be designed to accommodate such counter weights, as necessary. These may be up to 2.5% of the total gas lift. Alternatively, the gondola and tail support method may be configured to allow movement before fixing their positions. This complication, untried so far, may be difficult to achieve in practice and will add weight for the adjustment facility. Excess structure weight may easily result if tolerance limits are not kept tight and strictly controlled. The structural designer must also scrutinise strength reserves to minimise the airframe’s weight, since this has a much larger penalty on an airship’s performance compared with that of heavier-than-air craft. Increasing the power of an airship’s engines is not a feasible way to overcome excess weight problems. This can only be achieved by rigorous weight control or, if things really get out of hand, by increasing the size of the gas containers or envelope. Because an airship’s structure is so relatively delicate its mass and large shear force inputs need to be spread over as wide an area as possible. Particularly inputs from the gondola, tail surfaces, bow structure and propulsion units. This prevents excessive deflections and enables an efficient structure to be designed. Inevitably. however. large concentrations of mass are bound to occur: particularly in any attempt to locate the airship’s various systems conveniently (for example, everything installed in or from the gondola) and maintain balance. Compromises reached in this area establish the nature of the eventual structure, its layout (rigid, semi-rigid, or non-rigid), and the installations possible. Lightning Strike and Electrical Bonding Philosophy
Lightning strike is an ever present danger with airships, since they operate at altitudes with medium to high risk for lightning conditions. A combination of policies may be adopted, as follows:
e
To protect rigorously all habitable, critical parts and systems, by electrically bonding each area via a common earth system and providing conduction paths leading to static discharge positions.
e
To reduce the potential for a lightning strike attachment on critical parts - such as the: envelope, bow structure, empennage, bracing cables, gondola suspension cables, handling and control lines - by making them electrically non-conductive.
e
Avoidance of regions of known adverse lightning conditions.
174
Airship Technology
The first method is the conventional heavier-than-air craft manufacturer’s approach, which, for such large structures (as with airships), increases the potential for a lightning attachment. This is due to the arrangement of bonding braids and structural parts that results from this approach, which offer numerous points with a high potential for a lightning attachment. Of course, if the airframe is predominantly metallic or is made from electrically conductive materials (such as carbon reinforced plastic) this method may be the only practical solution. Rigids built in the past needed this method owing to their vast metallic skeletal frames and long bracing or control wires. Critical structural parts and control lines should be made with sections able to carry the resultant charge without failure. The second method is preferable, since it reduces the potential in the first instance. When lightning does strike it is then much more likely to make an attachment on a part with high potential, such as the gondola (which must have the first method applied to protect the occupants and avionic systems, etc), ignoring the rest of the airframe. Where electrical cables must be run (such as for navigation lights or power of control systems), these should be backed by earth braids capable of carrying the lightning charge and, where they run next to critical structures (such as a non-rigid’s envelope), backed with insulation material to prevent burn through. The potential for lightning conditions to develop is ever present and a pilot may inadvertently encounter these. Whilst the third method should be employed, to increase overall safety, it should be used in combination with the other methods to reduce the risk. Factors of Safety
Generally, where materials and methods of construction are similar to those for heavier-than-air craft, factors of safety that must be applied in the design of the airframe are common. These largely depend on the certifying authority but can vary, depending on national standards and whether the airship is to be regulated under civilian or military codes of practice. The second and third references are typical civil authority national requirements for the UK and USA respectively, which have been developed and are now being utilised to help formulate the basic standards for the rest
of the world. These standards define the various safety factors and other requirements
that must be applied to the design of an airship. In general, the authorities require that the maximum applied external (limit) loads be determined for critical flight or ground manoeuvres prescribed by them. The limit
load is “the maximum load anticipated in normal conditions of operation” (second
reference). These external loads must be placed into balance against inertia, including virtual inertia - which for an airship has a very significant effect - as appropriate. The
structure’s internal loads should then be determined, using a reliable approach (with
known confidence limits), which does not underestimate the results. The material's appropriate proof (yield) and ultimate factors (typically 1.0 and 1.5 respectively for metallics - but beware of higher prescribed values on castings, cables, etc) must be
applied in combination with other prescribed factors (such as fitting and form factors) as necessary. The resultant internal loads should then be compared with agreed
Structures
175
allowable loads for the material’s proof (yield) and ultimate (breaking) strengths. to ensure these are not exceeded. For composite and other non-metallic materials, or where doubt may exist, the appropriate factors of safety should be determined in consultation with the relevant certifying authority. This is necessary to ensure an adequate level of strength remains throughout the airship’s operational life. allowing for degradation in service. For fabric structures particular to airship technology, an ultimate factor of 4 is usually required. This is necessary to account for uncertainties in the basic material's and assembled fabric structure’s manufacture, degradation in service and creep rupture effects. It is also necessary to keep the material’s residual stress level low, in
order to avoid the possibility that an accidental tear might propagate rapidly (there being very few natural rip stopping members in a fabric structure).
In this respect, the basic material’s tear strength must be maximised and should not be degraded owing to age or by service conditions (although it usually increases with usage). It is usually necessary to demonstrate the tear strength by test using large scale representative methods, as well as standard small pull tests - particularly where large volumes of pressurised gasses are contained - to prove that tears will not propagate. Prescribed factors of safety (ultimate accelerations) on passenger seats, harnesses and the attachment fittings for general equipment installations vary markedly at this time between the different national airworthiness authorities. They are, however, significantly lower than for heavier-than-air craft, owing to the benign aspects of an airship’s flight. Current ultimate acceleration factors (g’s) for the UK and USA are given below (BCAR and US ADC): Table 7.1.
Engines & Propellers
Current ultimate acceleration factor (g's)
176
Airship Technology
Hopefully these differences will be levelled in the near future. In addition a 1.33 safety factor is required on local attachment fittings for seats and harnesses. Finite Elements
Finite element (FE) methods may usefully be employed to resolve redundancies and determine the internal loads, deflections, vibration characteristics, and the like,
for an airship’s conventional and essentially rigid structures, as paralleled within the heavier-than-air craft industry. These methods are now fairly common and can be studied in numerous literary works, so they will not be covered here. Special techniques and proprietary software, however, are needed in order to be able to use the method for fabric structures. To analyse fabric structures properly and fully, using FE methods, the proprietary software must be capable of solving large displacement, non-linear problems. It must also have membrane elements able to take up complex double curvature shapes, that have bi-directional properties dependent on the loads applied, which only operate in tension and with minimal shear capability. In addition it must be possible to create the model from a notional idealised geometric shape, give it an initial pre-tension (modifying that shape, simulating pressure stabilised conditions) and then load it in stages to add: gravity, flight accelerations, external aerodynamic and internal gas pressure distributions, point load inputs, and the like - each loading stage creating new shapes and boundary conditions. Where fabric structures interact with essentially rigid structures, the software also must be able to handle the differences and provide methods (elements) which can be used to connect the two types of structure. These aspects are necessary since, until a fabric structure has been pre-loaded (pressurised) it cannot act in its intended role and does not have a definable shape being just a heap of crumpled material on the ground. When stabilised by pretensioning (pressurisation) a fabric structure takes up its working shape. which is different from the idealised shape used for its initial design and manufacture. Successive load inputs then modify this shape. These shapes are created by initially running an idealised model with stabilising pre-loads to produce output that can be used as input for a new model, upon which the other loads can be applied - the procedure being repeated for each load step.
A basic assumption of the above is that the assembled fabric structure initially conforms with the idealised shape. This is not true if it has complex double curvature (as for envelopes and ballonets), since the fabric gores are cut from flat strips. Approximations, difficult to model, inevitably arise in order to construct the item. When pre-tensioned (pressurised) local stresses are created in the fabric membrane
owing to such approximate methods. It is assumed that these are negligible if the fabric material has compliant properties and the number of gores used in the assembly is optimised to reduce the anomalies to a minimum. Clearly, validation of such FE models is a necessary step, to ensure that the results
generated are within acceptable limits before using them in the design process.
Physical tests on representative fabric structures, which are also modelled using the FE methods intended for the design process, should be undertaken to validate the
Structures
177
modelling techniques and proprietary software's ability to predict the results accurately. Following such tests comparisons with the results of test and FE models should be made to establish confidence and use limitations. Having established confidence in the FE method, models of simple fabric structures may be created, which are easily validated by standard methods - gradually building confidence for larger more complex models. The remaining part of this chapter concentrates on structural designs for non-rigid applications.
PRINCIPAL STRUCTURAL GROUPS The primary structures of a non-rigid airship may be grouped under the following headings: envelope or hull; gondola or car; gondola support system: tail group; nose structures; landing gear; propulsion structures. Except for the last item (propulsion structures), the design methods of which are common with other aircraft and can be studied in numerous literary works, the design of each group will be discussed in the sections that follow. Envelope or Hull
Most of the recent non-rigid, helium filled airship envelopes, have been made from a coated fabric and/or laminated fabric/film material produced in roll form. These materials have been designed to contain the lifting gas under the loads imposed by flight and to endure the extreme environmental conditions of the airship’s operational regions. The basic material is cut and adhesively jointed to make a bag for gas containment, which is pressurised with air to stabilise and maintain its aerodynamic body of ‘revolution’ shape. The finished envelope is normally designed - using a single membrane - to contain the lifting gas, with integral internal ballonets (sacks or bladders) for the pressurising air. As an alternative to this method some airships have been designed using an outer structural fabric envelope, pressurised with air, with a separate inner bladder of film material, similarly shaped, to contain the lifting gas. The ballonets are made from a similar but lighter, thinner material to contain and keep the air from mixing with the lifting gas within the envelope. They are filled via an external supply system to pressurise the envelope. The envelope’s super pressure (i.e., the pressure at the lowest element of the largest cross section) must be maintained to prevent the envelope from loosing rigidity (becoming limp) - as height, temperature or atmospheric pressure changes - by allowing the proportion of ballonet air to vary, compensating for the volumetric changes of the lifting gas. Thus air (which is freely available) instead of the lifting
gas, may be added or exhausted to maintain constant envelope pressure. The ballonets may either be separate bags within the envelope, restrained by a cable system to load patches, or be integral with the envelope via an upper membrane continuously attached around its perimeter; with the envelope for its lower surface. to
178
Airship Technology
form individual airtight compartments. The first method may appear to be attractive from a maintenance point of view (since they are easily replaced) however, they are heavier (the envelope is there anyway). Separate ballonets are also more difficult to restrain if unwanted movement (slosh) occurs and the surface area inside the helium space is doubled increasing leakage rates. The ballonets are normally located against the envelope’s lower surface, since air being much heavier than the lifting gas - tries to find the lowest level (like water mixed with oil). When designing the ballonets this effect must be considered in order to prevent ballonet slosh, otherwise, as the airship pitches, large movements changing the airship’s trim may occur. In a dive, if the ballonet air sloshes forward - trying to find the lowest level - then the lifting gas will move aft to fill the void, causing the centre of buoyancy to move aft. If not kept within reasonable limits this effect could compromise the pilot’s ability to right the airship, tending to steepen the dive. If the ballonets are not properly shaped then some form of restraint will be necessary to prevent such slosh, increasing the weight of the system. The ballonets’ positioning on the lower surface is of benefit when considering the effects of damage from projectiles in minimising the potential for lifting gas leakage. The ballonets should also be positioned to avoid slosh against internal cables (such as gondola suspension) and other components within the envelope, to avoid abrasion damage. The air supply trunks should also be as short and straight as possible to minimise system losses. The size of the ballonets is dependent on the maximum altitude the airship is expected to fly at and the ambient temperature of the local region. Typically, they may be expected to have a maximum capacity of around 25 to 40 percent of the envelope’s gross volume, although they must be designed to collapse totally without straining the envelope at the pressure altitude (the altitude at which the lifting gas completely fills the envelope’s volume). If integral ballonets are fitted then additional slack must be provided in the ballonet material to allow for the envelope’s expansion under load (the ballonets being pressed against the envelope and forced to adopt its profile), otherwise they will restrain the envelope causing stress concentrations and picking up load themselves. The number of ballonets may range from one to four. For a small airship one ballonet may be considered sufficient, if there is no interference with the gondola’s
suspension system. Medium or large airships of low altitude can take advantage from
siting two ballonets, one forward the other aft; thus, by shifting air between them
(allowing one to inflate while the other collapses) the centre of buoyancy of the lifting gas can be moved to effect a change in trim. This can only be done near to the ground, since at altitude the ballonets are nearly empty; thus enabling the airship’s balance to be adjusted for landing or take-off when aerodynamic control may be poor owing to low flight speed. For an airship of high altitude needs and/or in hot climates a large ballonet capacity may be necessary. In this instance the additional capacity can be provided via pannier ballonets each side. The pressurisation system of a non-rigid airship is very important, since this is what
maintains the envelope’s ability to perform as a structure. A dual system, with independent trunks, fans and controls, etc, should therefore be considered, to ensure
Structures
179
the envelope’s working pressure is always maintained. In addition a means to pump air directly into the lifting gas area is needed to pressurise the envelope under emergency conditions. This may occur if the envelope is holed (for whatever reason). enabling the lifting gas to escape. Under such conditions the ballonets inflate until they are full. When full they can no longer expand freely to pressurise the envelope. and thus become ineffective. Under normal conditions the differential pressure across the ballonet membrane is very small, reaching a maximum at their lowest point (where the head of air is greatest). This is particularly important for the design of pannier ballonets. Slosh increases the pressure, owing to the dynamic effects that occur when the moving mass of air is stopped by the ballonet’s membrane. The ballonets’ design pressure, however, is normally established from their full condition, since, if the fans are left running with the valves locked and closed, they merely serve to pressurise, tending to burst the ballonets. Because of the power of the fans (required to overcome resistance along the trunks and deliver air to the ballonets) the pressure developed by them is usually an order of magnitude greater than normal operational conditions within the ballonets and envelope. Care must be exercised in the hangar to avoid rupture of the ballonets if, for maintenance purposes, other, more powerful, fans are used to maintain the airship’s pressure. The air supply rate to the ballonets must be designed to maintain constant envelope pressure under the airship’s maximum rate of descent and/or expected extremes of climatic change, compatible with the volume of the envelope. This invariably requires an air supply system with powerful fans to cope with the vast quantities of air that need to be moved. Tapping air from behind the airship’s main propellers, using scoops located in the propeller wash, and then ducting it via trunks is the most convenient way of supplying the air under pressure to fill the ballonets. Auxiliary fans, however, are also needed to deliver air when the propellers are not functioning, vectored out of line from the scoops, in reverse thrust mode or at low power. Retractable scoops may also be fitted so that propeller efficiency is not degraded when the envelope’s pressure is stable (which should be most of the time). If more than one ballonet is fitted, a distribution system is also needed to direct the air to particular ballonets. In order to reduce the volume of entrapped air within the ballonets, large valves must be installed to release the air under the envelope’s pressure into the surrounding atmosphere. This is necessary in order to maintain a constant envelope pressure when ascending or owing to increases in pressure from changes in weather. These valves must open automatically as well as manually and be able to operate at different pressure settings, as conditions dictate, to regulate the envelope’s pressure. High envelope pressures are needed for turbulent weather and rapid manoeuvres, whilst lower settings may be adopted for hangar use and calm conditions. The ability to lock valves with low settings enables higher operational pressures to be maintained. Similar vent valves are needed to relieve envelope pressure directly, by releasing the
lifting gas to atmosphere. These safety relief valves need to operate automatically
after the ballonets have completely deflated when pressure is rising above acceptable limits for the envelope. The relief valves also require manual override to test their
180
Airship Technology
function, which may also be used to enable a balloon landing or flight control under emergency conditions. The gas pressure relief valves should be located at positions that enable the lifting gas to be released efficiently but without the possibility that all of the gas could escape, should they leak or fail to close (which could occur if they are sited at the top of the envelope). This is a danger, since they are only operated infrequently with the consequent danger that they might seize or stiffen through lack of use. The designer should therefore pay particular attention to the valves’ details to ensure reliability in service. Since the flow rate through these relief valves may be high, to prevent the danger of losing too much gas, should they fail to close (which not only reduces the buoyancy but also compromises the ability to maintain envelope rigidity - owing to the ballonets’ capacity, when full, to compensate for the lost gas volume), small gas vent
valves may be installed for control purposes under conditions of free-balloon flight. These valves should be designed so that if they do fail open or leak, the rate of flow through them is small, such that continued flight is prolonged, making a safe landing possible. Inevitably the various valves and fittings require penetrations in the envelope. The envelope therefore must be reinforced to enable the fabric tensions and resultant stress concentrations to be carried around the resultant holes. In addition, owing to hoop and longitudinal stress differences, the envelope’s fabric tensions are variable around the penetration. A stiff edging ring is therefore also needed to maintain the hole shape. Since the valves are fitted to these penetration positions they must be able to accommodate resultant distortions without affecting their operation and performance. The continual supply of air to and release from the ballonets, causes them to rise and fall frequently. This flexing, together with flexure owing to slosh, places a heavy duty on the material and its joints - particularly, for integral ballonets, at their intersection with the envelope. Materials with a high flexural and gas retention capability are therefore needed to overcome associated problems. The additional materials of the
ballonets’ joints with the envelope also introduces stress concentrations, owing to the inevitable increase in envelope stiffness at these positions. Orientation of material and graduation of reinforcing plies or jointing tapes should therefore be optimised to reduce the stress concentrations to a minimum. To calculate the pressure required for maintenance of envelope rigidity in flight the maximum applied bending moment must be established. To do this the aerodynamic loads on the envelope and empennage must be put into balance against the mass distribution (including concentrated loads from the tail, gondola and nose structures) subject to gravity and the accelerations resulting from flight manoeuvres or gusts, buoyancy (not forgetting to subtract the ballonet effects), thrust and drag, etc, and then integrated along the length to determine the maximum value. The lifting gas static head effect (pressure gradient from bottom to top) also must be taken into
account. Empirical formulae for the determination of the maximum bending moment in flight, as a short cut to the above, have been used in the past (Burgess, 1927), The FAA (US
Structures
181
ADC) also provide a formula, as given below, which may be used by the designer in the absence of a more rational analysis. It is recommended that these formulae only be used as a check, to validate a dynamic analysis of the airship’s flight envelope (from which the maximum bending moments may be derived). FAA bending moment formula:
M = 0.029{1 + [L/d - 4][0.5624L9-02 - 0.5]}p.u.v.V.L9-25
(7.1)
Applicable for L/d between 4 and 6. For L/d < 4, use 4.
Where: L
=
Length of airship (ft).
d
=
Maximum envelope diameter (ft).
p
= _ Density of air (slugs/cu.ft).
u
=
v
= _ Airship equivalent speed (ft/s).
V
= _ Total envelope volume (cu.ft).
Gust velocity (ft/s).
Note: Gusts of 25 ft/s, while flying at the maximum level flight speed at sea level, and 35 ft/s, while flying at the design speed for maximum gust intensity should be considered. To maintain rigidity the envelope must remain under positive tension when subject to the maximum bending moment combined with other compressive load inputs. Assuming that engineers’ theory of bending is valid and the envelope’s cross section is a circle of radius ‘r’ one may write: f).t = M.r.(t/I) + (T - p.(1.r2))/(2. 2.1) > or = 0
for rigidity. Where: f;
=
Longitudinal membrane stress.
t
=
Membrane thickness.
M
=~
Maximum bending moment, including pressure gradient effect.
I
=
Membrane second moment of area.
T
=
Thrust.
=
Envelope super pressure.
C772)
182
Airship Technology
ew,
§=
3.142.
For a circular membrane of large radius it can be shown that:
Vt = 1.13
(7:3)
The minimum pressure (pin), Where rigidity is just maintained (without a kink
developing), may be calculated from the condition for f,.t = 0. Re-arranging and substituting for I/t we can therefore write:
Pmin = 2-M/(n.r3) + T/(xt.12)
(7.4)
Neglecting thrust and re-arranging, the bending moment ‘M,’ which just overcomes rigidity (kinking the envelope) may be defined as:
M, = (p.1.13)/2
(7.5)
Note: engineers’ theory of bending has been shown by test to be valid for non-rigid airship envelopes to a reasonable degree of accuracy, so the above formula is valid (Burgess, 1927).
Having determined the pressure, material for the envelope’s manufacture may be established by treating the envelope as a cylinder and applying pressure vessel theory, where:
Hoop tension, f}.t = p.r
Note: When dealing with fabrics membrane stress is a meaningless term, since fabrics are made from individual fibres instead of a homogeneous material. Thickness, if used, is an idealised term - which may be necessary for certain computing techniques. The term ‘f.t’ is therefore used instead, being the load or tension per unit width, which can be determined by ordinary test techniques. If the bending loads overcome rigidity then the envelope will collapse, a kink developing as shown in Figure 7.2. When, however, the load is removed the envelope normally pops back into shape without any damage and is then able to perform as a rigid structure again. This phenomenon is very useful if accidental overload occurs, since the structure suffers no permanent damage, enabling the airship to continue flying. How many other aircraft, after being subjected to such overloads, could continue as normal? Clearly this behaviour warrants further investigation. When the envelope kinks a wrinkle develops around the cross section from the point where the membrane tension is reduced below zero by the bending effects, as shown in Figure 7.3 (without breaking or releasing the lifting gas), until a stable position 0 is reached where the tension becomes positive again.
Structures
Kink
Envelope
Figure 7.2.
Kinked envelope owing to lack of rigidity.
Kinked
portion no
longer effective
4
Figure 7.3.
Section still effective
Extent of kink around the cross section.
183
184
Airship Technology
Assuming that engineers’ theory still holds true and the cross section remains circular, the effect may be treated as follows: the wrinkled portion of the envelope is no longer effective, so it should be removed from calculation of the section constants. Revised section properties may be established from:
Vt = r3.[ - 0 + (sin2{m -0})/2 - 2(sin2{™ - 9})/(x - 8)]
(7.6)
and the position of the section’s neutral axis is:
NA = r(sin@)/(z - 8)
(7.7)
below the section’s centre.
The centre of pressure, however, remains at the section’s centre. Neglecting thrust, for equilibrium we may therefore write: P/(2.r.t. {70 - 8}) + P.NA.(r.cos® + NA)/I = M.(r.cos8 + NA)/I
(7.8)
Where: P = p.m.r2 Hence, re-arranging and substituting for P we can therefore write:
Mi(xpr°)
=
|
m—0+cos 6 —sin@ 2 [sin + (x — 6)cos }
= k
|
(7.9)
From a Plot of 8 against k, as shown in Figure 7.4, we may easily determine bending moments demonstrating that, while the envelope may have kinked it is still able to resist bending. In fact, when: 6 = 180 degrees, k= 1
(7.10)
Therefore:
M = p.z.r3 = 2.M,
FARE
Clearly, an envelope that has kinked to such an extent is not feasible, the theory
having broken down owing to non-linear effects. To examine the theory further the longitudinal tension (f3.t) at the envelope’s opposite face will be examined.
Structures
whenO9=0°
-:k=0.5
then M=M,=0.5pAr
when 6 = 180°
:k=1.0
then M = pAr = 2M,
where
a
185
M, = Wrinkling moment
n-0+ cos 6-—sin® 2 [sin® + (x -@)cos ‘|
LTT ATT RRR RRRREDZARRREEE TT eT mapa a cae Internal pressure: p
0
20
40
80
100
120
140
160
180
8 (Degrees) ———
Figure 7.4.
Graph of 0 against k - Theoretical buckling of flexible cylinder under internal pressure due to applied bending moment.
186
Airship Technology
Similar to the above, for equilibrium we may write:
f) = M.(r - NA)/I + P/(2.1.t.{0 - 8}) - P.NA.(r - NA)/I
(7.12)
Also, from above:
(TASS
M=P.r.k Hence, substituting for P, I, NA, M and k, and re-arranging we get:
f)./(p.1.r) = [(1 + cos0)({ - 0} - ({x - 8}.cos8.sin8) -
2sin20))/[2({n - 0}.cosO({n - 0}? - (3sin2®) + {7 - 0}? - 2)))] (sin39(
=K
(7.14)
From a plot of 6 against K, as shown in Figure 7.5, we may easily determine the longitudinal tension. However, as the Figure demonstrates, beyond about 110 degrees the tensions rise by an inordinate amount - approaching infinity at 180 degrees. This cannot be, since either the material would have failed (unlikely) or other non-linear effects would have taken over, the original assumptions no longer being valid. When,
60=0,M=M, and K = I/n
Therefore,
f,.t=p.r=f;.t
When,
6 = 110 degrees, K = 0.63 = 2/n
Therefore,
f,.t = 2.p.r = 2.f).t
The above is offered to raise awareness of the potential benefits and effects of pressure stabilised envelopes, but does not purport to be a complete theoretical
approach. Testing to determine the limits for which it may be valid needs to be done before relying on values determined from it. With the availability of powerful nonlinear large displacement finite element tools, able to handle fabric structures, a more accurate approach might be developed. The envelope may be manufactured from strips of material (known as gores) either running longitudinally, like orange peel strips, or circumferentially - both techniques having been used previously. Whichever method is adopted, approximate methods need to be employed to determine the gore’s shape for cutting, since the material is manufactured flat and then is forced by internal pressure to adopt a shape with double curvature. Clearly, this introduces stress concentrations (straining some areas more than others), which should be kept to a minimum.
Structures
RURRERERERERR EDEL (a)
(b), , (c)
4-0
Internal pressure:p
|
Le.
(d)
Area:A= nr? 2d Mom. Area:!
(a) = M(r-y) (b) = nrp/2r(n-6)
(c)= arpy (¥+rcos6)/I
(d) = [nr’py (r-y)VI K=trf,/AP When
6=0,
M=M, = O.5pAr,
K = 1/n = 0.3183 then
aie) 0-6
f, = Ap/nrt = 2M,/At
2 PCH ALL cereale SeCesee
jasaeGauReae= deeg ae és SS ep EER EEE RE Sa ah i Se nt 60
80
100
120
140
160
180
@ (Degrees) ——=
Figure 7.5.
Graph of 0 against K - Theoretical maximum tensile stress of inflated tube in port wrinkling regime.
187
188
Airship Technology
A compliant fabric should therefore be used. When determining the gore’s width (apart from restrictions owing to available supplier widths) a balance needs to be
struck between the number of joints (which should be minimised) and the strain induced (which reduces as the number of gores is increased). The seams should be designed to develop the full strength of the fabric being joined, and be gas tight (unless a separate gas bladder is employed). Modern fabrics may be laminated with films (to provide a gas seal), which usually are difficult to bond to and have poor peel strengths from their base fabric. As a result it may only be possible to join the envelope material from one side. Modern adhesives, however, can now be used to make the joints without stitching; provided a comprehensive test programme has been undertaken, which demonstrates the various joints integrity under all conditions of service. Quality control and cleanliness during manufacture is of the utmost importance, to ensure the actual joints meet the required standard. A typical joint may be made by butting adjacent gores together with a small gap (about .030”) between them, to prevent a ridge from forming when joined. The joint is made using a structural tape bonded on the strength side, the side of the material which produces the best joint (usually, for a laminated material, the opposite side to the gas retention film). In addition, a light non-structural tape employing film material on the opposite side may be necessary to restore the gas seal. The structural tape may use material of similar construction to the basic envelope, but without the film material. The various jointing tapes should be of material cut on the bias (+/- 45 degree), to enable the joint to expand with the envelope. In addition, the jointing tapes (structural and gas seal) should be of differing widths to reduce stress concentrations at their edges. Longitudinal gores have their fabric’s weft threads, which weave in between the more straight warp threads, running circumferentially. If the envelope is treated using cylinder theory then the circumferential or hoop tensions are twice the longitudinal tensions. As a result, when the envelope is pressurised these threads try to straighten against the warp threads (which are bent by this action) as well as stretch under load. The effect of this is that, when pressure is applied, overall length changes may be small while the diameter may enlarge significantly. If the gores are circumferential then both longitudinal and diametrical changes may be expected. These effects must be taken into account when determining the envelope’s final volume and the position of components on and interfaces with the envelope. Assuming the envelope to be a body of revolution, with form derived from a polynomial expression y = f(x), its geometrical attributes may be determined from the following standard mathematical formulae: Volume = n.Jy2dx
(7.15)
Prismatic Coefficient = n.r2.length/Volume
(7.16)
Centre of buoyancy = (n.Jx.y2dx)/V olume
(7.17)
Surface Area = 2.nfy.({ 1+ [dy/dx]2}(12))dx
(7.18)
Structures
189
Centre of gravity = (2.n)x.y.({1+[dy/dx]2}U/2))dx)/Surf Area
(7.19)
Longitudinal surface length = J({1+[dy/dx]?}[!/2])dx
(7.20)
For circumferential gores, Blakemore (1927) gives a method to define the shapes to be cut. For longitudinal gores a simple patterning method may be established, treating the envelope as a body of revolution, from the following procedure: e
Divide the envelope along its axis into a number of stations (1).
e
For each station (xi) calculate the radius (yi) of the section from the polynomial expression y = f(x) for the envelope’s shape.
e
For each station (xi) calculate the longitudinal length (si) along the surface to the station using the expression above.
e
For each station (xi) calculate the gore’s half width from the expression wi = (1 /n).yi.(1-{p.yi/{E.t]}), where n is the number of gores and [E.t] is the fabrics modulus.
e
Plot si against + and - wi to define the gores shape.
Note:
e
The term (n/n).yi2.p/[E.t] is a reduction term to allow for the envelope’s expansion under internal pressure, otherwise the volume (and hence buoyancy) would be too much.
e
For an envelope shape with rounded ends the expression for longitudinal surface length cannot be calculated, owing to singularities at these positions. As a result approximate methods at each end must be used to overcome the problem. Gondola or Car
The gondola or car is an aerodynamically shaped vessel, similar in principle to and the fuselage of heavier-than-air craft, provided to house the crew, passengers from arise may gondola term the airship’s general systems. (Chamberlain, 1984), some early Zeppelin terminology owing to their boat-like appearance. Certainly, previous airship gondolas have been used to enable airships to settle on water (Williams, 1974); an attribute that may be deemed highly desirable.
A single gondola has been the norm for non-rigid airships, although separate load over the nacelles may be used to help spread the subsequent concentrated advantage of this envelope’s lower surface. Rigid airships of the past have taken
190
Airship Technology
option, providing separate engine or power cars. Separate cars enable noise and vibration to be isolated from the habitable areas and provide a wide disposition for propeller and undercarriage units. However, access in flight needs to be provided to each nacelle if long endurance flights are envisioned; balance is more difficult to maintain as fuel is used or other variable loads are disposed; cabin heat, power
distribution
systems
and
lightning
strike
protection
is more
difficult
to co-
ordinate/devise. Placing as many systems as possible into a single gondola is the most convenient option and, from a weight/drag aspect, is more efficient, although the suspended mass is difficult to carry without noticeable envelope distortions. As airship size increases the proportion of gondola suspended mass to the overall airship mass increases, the gondola becoming more of a burden to support. At the same time the envelope’s ability to support the suspended mass without significant distortions occurring diminishes with increased size. In simple terms, this is because the envelope’s ability to support the gondola’s suspended mass is proportional to fabric tension (i.e., proportional to ‘pxr’), whilst the gondola suspended mass is roughly proportional to gas lift (which in turn is proportional to envelope volume, i.e., proportional to ‘r3’). So, if ‘r’ is doubled the suspended mass is increased by a factor of 8. At the new size the pressure required to maintain rigidity increases by a factor less than 2 (mainly due to higher speeds). The envelope’s ability to support the gondola’s suspended mass therefore only increases by a factor of 4. Overall, the envelope’s ability to support the suspended mass is therefore diminished by a factor of 2. As a result the gondola’s length becomes increasingly important, to be able to spread the load over a sufficient length of the envelope. Since the gondola’s depth is roughly controlled by human proportions (the same for any sized airship) its length to depth ratio rises, reducing rigidity and increasing bending stress levels. This can be offset by making the gondola double decked, but at the expense of concentrating mass (making the suspension system more difficult to design). Clearly, compromises are necessary as size increases to optimise the airship’s overall design.
Depending on overall configuration, the gondola may be designed to carry the propulsion and power units, fuel, ballast, water recovery system (if used), landing gear, electrical systems, avionics, envelope air supply, furnishings, general equipment, mission and other systems of the airship. As such it is the working centre of the
airship and so should be designed for a variable interior arrangement. It should be located such that when fully laden the suspended weight centre of gravity (cg) is roughly below the lifting gas centre of buoyancy (cb). Variable masses (fuel, ballast, crew and passengers, etc) should be located within the gondola to maintain overall airship balance.
The type of gondola provided depends on whether cabin pressurisation is required
for the crew and passengers at altitude, since this has a dominating influence on the structure. If pressurisation is required then a car of circular form is needed to optimise
the structure as a pressure vessel. This shape is awkward to work with, since it does not conform to the envelope’s profile (also of circular form), making the interface complicated and the support system difficult to configure. Inevitably there is also a
Structures
191
weight penalty, since the structure has to be stronger with increased complication/parts count and more difficult utilisation of space compared with an
equivalent non-pressurised car. In an attempt to ease the problems of envelope interface and space utilisation the pressurised car may be shaped as shown in Figure 7.6a. It must be recognised, however, that as a pressure vessel this shape is not ideal. This is because it only partially works in membrane tension - imposing greater bending loads on the primary members and joints to hold the shape as it tries to become circular. Optimisation exercises looking at aspects of overall airship design - not just structural - need to be run to determine the best shape. For a non-pressurised gondola much more simple shapes may be adopted, extending up to nest against the envelope, as shown in Figure 7.6b.
Vertical
suspension
Sway brace
Vertical suspension
Envelope
Envelope
Collar
a) Pressurised
Figure 7.6.
Gondola cross sections.
b) Non- pressurised
192
Airship Technology
The type of gondola structure is heavily dependent on the airship’s role. Examples of different roles and how they affect the gondola’s design are given below: Ex. 1
Small airship for pilot training. The gondola of such an airship may be of simple construction, essentially a faired bedstead frame with cockpit to mount the necessary systems and accommodate the instructor with his trainee. No passenger accommodation is necessary. Thunder and Colt’s GA-42 may be used to illustrate this category.
Ex.
Medium sized airship for sightseeing purposes. The gondola for this role should be configured with a comfortable sized cabin, with large panoramic windows, to accommodate a passenger seating arrangement with galleys and toilets, etc. Standard seat tracks should be fitted to allow variability between customers. To aid the passengers’ sense of well being, the structure should
have a good appearance with a feeling of security about it. It should also be able to handle general abuse. Airship Industries SKS 600 series serves to illustrate this category. Ex. 3
Medium sized airship for para-military surveillance duties. This is similar to Ex 2 but not so fussy on creature comforts. Seat tracks are still required, but are used more to mount customers equipment. Additional hard points may also be needed to mount special items, such as radar antennae, camera pods,
search light, rescue winch, etc. Also, extra apertures may be necessary and doors widened to accommodate special needs. Westinghouse Airship Inc’s
Sentinel 1000 illustrates features for this type of airship. Ex. 4
Large oceanic airship for patrol purposes. This may be a scaled up version of Ex 2 and 3, which also should provide: hotel-like accommodation for extended length flights, large capacity fuel and ballast tanks, a cargo deck with large floor aperture and winches for inflight replenishment. The projected design for the US ODM programme, designated YEZ-2A is an example of such a design.
Standard structural techniques, taken from the heavier-than-air craft industry, should be employed for the gondola’s design. These can be studied in various literary works
on the subject: Bruhn (1973), Peery and Azar (1982) and Niu, (1988) being just a selected few.
Gondola Support System
Early designs (Abbott, 1989) adopted an external method of support from finger patches on the envelope’s lower surface with the gondola suspended below,
using cables. This method had a number of merits, such as:
Structures
¢
193
Simple concept - easy to install rig and maintain - that does not require envelope penetrations or entry, which otherwise would require breathing
apparatus and cause additional complication. e
Easy to remove/install or switch gondolas, enabling modifications undertaken in convenient workshops or new marks to be fitted.
¢
Substantially reduces landing load inputs, since the gondola is separated from the envelope and the suspension cables can go slack, preventing load transmission from the envelope.
to be
In order to spread the gondola’s suspended weight over a sufficient area of the envelope’s lower surface and provide an essentially tangential input, the gondola was generally underslung some way below. This introduced disadvantages, to be weighed against the above, as follows: e
Overall airship height increased, requiring taller hangars and masts.
e
Makes interfaces for necessary controls and systems difficult to configure.
e
Increases drag (lowering overall airship performance) owing to greater gondola wetted surface area, gondola outside the envelope’s boundary layer, large total
length of cable across the airflow and interference effects between the envelope and gondola.
e
Limits permissible pitch attitudes owing to cables going slack.
e
Lowers overall airship centre of gravity, increasing pendulum stability, making the airship more difficult to manoeuvre. As a result larger (more powerful) control surfaces are needed to overcome the effect.
e
The gondola cannot easily be used to mount large mission system antennae - a feature which may be desirable for current surveillance duty objectives.
Current airships have sought to redress these problems by configuring the airship with a close-coupled support system, enabling the gondola and envelope to nest together. Some airships have even placed the gondola (ZPG-3W, designed by Goodyear in the 1950s) partially inside the envelope via a large aperture. The
resulting configuration largely depends on the airship’s role, size and gondola section profile. For small airships (60,000 cu ft - 1,700 m3) configured for light aerial duties (advertising and/or camera platform) of short endurance (8 hours) the gondola’s suspended mass is a small proportion of the overall airship mass, the fabric tensions
are proportionately large, and limitations on all weather applications may be imposed (compared with medium or large airships). As a result a simple external support system may be considered, as utilised by the American Blimp Corporation's A-60 Lightship. Larger airships require a disproportionately bigger footprint to carry their gondola’s suspended mass, which is difficult to provide on the lower surface. An internal
194
Airship Technology
suspension system, supporting the gondola from the envelope’s upper surface, may be used to overcome the footprint problem, since it uses the envelope’s large diameter to allow the load to be spread over a wide area. If the cable splay is limited to about 15 degrees from vertical then the resultant compression over the envelope’s upper surface will not overcome the membrane tension. Essential features of a typical close coupled support system are:
e
Internal suspension cable system, which essentially reacts vertical weight.
to the gondola’s
e
Gondola to envelope lower surface interface connections, which react to the gondola’s horizontal shear forces.
Load curtain
Bolt
rope
'D' Ring fitting
Suspension
cable
Forward
ballonet
Att ballonet
Suspension cable Envelope
Gaiter Clamp ring
Gondola
Bung
Channel fitting
Figure 7.7.
Typical close coupled support system.
Structures
195
As shown in Figure 7.7 the suspension cables are connected to fittings along the gondola’s upper sidewalls. They are then passed through gaiters in the envelope’s lower surface, which accommodate interface movements and seal the envelope. The cables then pass on up to arched load curtains, which in turn transfer the load to the envelope’s upper surface. The arched load curtains are longitudinal fabric structures hanging from the upper envelope, essentially acting like inverted suspension bridges, which should be designed to spread the cable loads evenly along their length. The input cable loads. which connect at the cusps, are fed into the curtains via continuous bolt ropes (load tapes) along their lower edges. These can also be backed by finger patches at the cusps, as necessary. Adjacent cables along the length tend to balance between themselves around the bolt rope, similar to a rope over a pulley. At each end. however, the unbalanced bolt rope load must be reacted by load patches into the envelope. The incoming cable should bisect the cusp angle equally. to prevent excessive load on one side of the cusp. In addition the included angle of each cusp should be made acute, to minimise the load in the bolt rope and prevent load entering the curtain too quickly. To assist the latter point, the bolt rope (load tape) should be of high modulus material so that the load remains in it over a longer length. The number of load curtains depends on the cable array, which may bifurcate or combine at different stages to help reduce the inputs at the attachment points. If more than one (two total) curtain for each side of the gondola are used then, in order to
keep the curtains as near perpendicular to the envelope’s surface as possible the cables must bifurcate and cross, as shown in Figure 7.8a for a four curtain system. To prevent the crossing cables from rubbing together the curtains should be staggered; the curtains on one side of the airship being displaced forward a small amcunt, whilst the curtains on the opposite side are displaced an equal amount rearward. Similar to a stretched line with a weight hanging from its mid point, the envelope cannot react to the vertical load in the curtain until it has deflected, forming a cusp as shown in Figure 7.8b. The resolved components of envelope tension in the opposite direction react the curtain’s load. This is where the simile ends, because (unlike the
stretched line) tension in the envelope does not rise as the deflection or load increases. since it does not have fixed end constraints. Tension in the envelope’s membrane is determined by pressure vessel considerations, so it continues to deflect (with essentially constant envelope tension) making the cusp more acute until a balance is reached. In order to avoid deep furrows in the envelope’s upper surface, the number of curtains should be increased as airship size is increased. Connections at the envelope’s lower surface primarily depend on:
e
The gondola’s sectional shape (see Figure 7.6).
e
Relative vertical position (large gap, nested or inserted).
e
Gondola suspended weight to length ratio, compared with the envelope’s shear carrying capability.
196
Airship Technology
a) General Cross Section
YY)
Ov
Load curtains
Upper suspension
cables
See detail below
Envelope P
Lower suspension cables
in all
membrane
Envelope
u
wz: Load curtain load per unit length
'
(f,.t)
(f;.t)
Load
curtain (f,. t) =p.r
And: we 2 (f;.t) cosa aescosty w :
b) Upper Envelope Joint
Figure 7.8.
Four curtain suspension system arrangement.
eee,
Structures
197
If the gondola nests against the envelope with a small gap (2.5 to 10 cm) and its
sidewalls extend up to the interface, then a simple shear collar may be all that is necessary.
This
is a strip of envelope
material
(about
20 cm
wide)
that wraps
completely around the gondola, being attached continuously at its lower and upper edges to the gondola and envelope respectively. Whilst the collar has no effective
shear stiffness, under load it buckles, allowing - provided its width is limited - a small movement at the interface until a diagonal tension field is set up (similar to a thin walled beam in bending), thus transferring the shear forces from the gondola to the envelope where they are reacted. Provided the shear flow (load/unit length) is within acceptable limits for the envelope (approximately 10% of the tensile strength, if wrinkling is to be avoided), then no further connections are necessary. A short gondola, however, may be effectively lengthened using battens (similar to those of the nose group) attached at the gondola’s extremities. In addition, a small portion of the shear collar at the front and back of the gondola helps to carry longitudinal loads by direct tension into the envelope. This only occurs at one end, where tension is created, since the opposite end goes slack. If the gap is large or the gondola’s shape does not permit a shear collar of small width, then movements at the interface may be unacceptably large for a shear collar alone. The gondola’s motion beneath the envelope may be compared with that of a four bar chain. Special methods for the particular application may need to be devised, incorporating cables from finger patches on the envelope (arranged in martingale fashion, to brace the gondola) and lower envelope surface battens to transfer the load into the envelope. Rigging methods to pre-tension the arrangements will also be necessary. In this application the shear collar may be more effective as a lower surface arched load curtain, supplementing the internal suspension system by carrying a proportion of the vertical suspended weight. It can then also act as a fillet fairing between the gondola and envelope, providing enhanced aerodynamic properties. In this respect the arches need to be closed by light weather-strips, which may be folded back to give access for maintenance purposes. If the gondola is partially inserted via a large aperture into the lower envelope surface then shears may be fed directly through the interface connections with minimal deflections. The aperture in the lower envelope surface should be designed using similar principles as for an arched load curtain (discussed above), using a continuous bolt rope from which connections with the gondola are made. The attachment points may also be supplemented with finger patches, as necessary. The connecting members should be length adjustable for rigging purposes. The gondola should be used to carry envelope tensions (dependent on gas pressure) from one side
of the aperture to the other. In addition a sealing membrane is required to close the
gap between the arches and the gondola. Weather-strips may also be required, as for the external load curtain described above. Whilst this last method may cause some consternation with respect to the size of the aperture in the envelope (as a pressure vessel), which adds to the complication When
the envelope is inflated, it should not be dismissed when considering the airship’s
198
Airship Technology
role. It has, after all, been done before, as evidenced by the US Navy’s ZPG 3W airship (42500 cu m). Merits of this method are as follows:
e
The large aperture may also be used to install large objects inside the envelope, such as mission antennae or additional fuel tanks.
e
Installations in the envelope could be connected directly on the gondola, as a stable base, without the need for additional envelope penetrations.
e
The myriad penetrations normally put in the envelope’s lower surface above the gondola for: suspension cables, viewing, access, ballonet air supply system, through fittings, etc, would be obviated.
e
Access into the envelope could be made safer and relatively simple, using a turret on top of the gondola with a sealing door and extra facilities for air supply. The turret could also be used to improve internal inspection.
e
The airship’s centre of gravity would be raised, improving handling qualities.
e
Depending on overall layout of the propellers, overall airship height could be reduced, thus easing hangar entry.
Whichever method of support is adopted, fail-safe design principles should be applied. Cable failures within the envelope are not easy to detect (particularly when one considers: the airship’s size, poor lighting conditions and difficult viewing facilities); it is essential that the system functions safely following such a failure.
Tail Group
The tail group normally comprises a set of fixed fins fitted with control surfaces from hinged brackets at the trailing edge. They are located at the airship’s rear end to provide stability in flight and, by operation of the movable trailing surfaces, effect control - similar in principle to the empennage of other aircraft. The fins may be held in place by an external support system. The fins’ actual position along the hull must be determined from an optimisation exercise between various influences as follows: ¢
The further aft the greater the moment arm, enabling small surfaces to be used, but read on.
¢
The further aft the smaller the hull’s diameter causing the surfaces to operate in poor air-flow behind the hull’s bluff fore body, which also causes a deep turbulent boundary layer, requiring larger surfaces of greater aspect ratio to overcome the effects.
Structures
199
e
The further aft the smaller the hull’s diameter and therefore the smaller its aerodynamic influence, which contributes significantly to overall empennage loads, requiring larger surfaces to compensate.
e
The further aft the smaller the hull’s diameter and therefore the smaller the envelope’s membrane tension and overall section, affecting its ability to support the fins, requiring a larger fin footprint, attachment fittings and possible stern reinforcement to spread the loads.
e
The further aft the more effect overweighi-fins will have, requiring larger counter balance weights to be added at the nose.
In addition to position, the number of tail surfaces (fin plus control surface), circumferential arrangement, aspect ratio, section profile and plan form must be determined. Having established these parameters the structural design may be undertaken. An objective of the design should be to make all tail structurally identical, to economise on their production. If an arrangement (such as cruciform), with the lower surface on the centre-line of symmetry, is selected the lower surface may have to be
of the units surface units
longitudinal truncated to provide ground clearance for take-off and landing. It may also require a fender to protect it from impacts with the ground. The arrangement should be configured so that snow will not settle on them or can be removed easily. A typical tail surface arrangement is illustrated in Figure 7.9.
Envelope patch
Fin
load
-Control surface
rcot
Figure 7.9.
Tail surface structural arrangement.
200
Airship Technology
The fin usually comprises a set of interlocked spars and ribs, sparsely arranged in lattice form to provide a stiff structure (similar to an egg box arrangement). As far as practical the arrangement should be squared to avoid complication. Depending on size, the fins may also need to be constructed from major sub-assemblies with joints that can be disassembled for transport purposes. Most of the lattice structure may be skinned with doped heat shrink fabric to provide a smooth aerodynamic profile. A NACA 0008 series profile of 8% depth to chord ratio, with its maximum thickness at 30% chord length, has been used in the past. Internal cross bracing to hold the lattice arrangement may also be necessary. The skin’s lower surface should also have small drain holes at positions where water may collect. The leading edge should be stiffened against aerodynamic pressure and effects of abrasion. This may be achieved using lightweight thin gauged sheet material, to form a D shape, backed by a raked spar. Similar sheet material should also be used between the root rib and the first rib outboard, to stiffen this area against the reactions from the envelope as the fin is pressed into it. The root rib should be profiled to conform with the envelope’s shape, be of smooth form and have generously curved edges and corners to prevent the envelope from being damaged. Fittings to install the fin on the envelope should be provided at the root end (for shear attachments) and at outboard positions (for bracing cables). In addition, fittings to enable the fin to be lifted into position may be necessary and hinge brackets, extending rearwards from the fin’s rear spar, are needed to carry the control surface. The control surface may be like a barn door, hinged at its leading edge. Such an arrangement, however, would require large aerodynamic horns and powerful tabs to help reduce the hinge moments and/or be fitted with a powerful operation system (which may be heavy and slow to operate). In addition, the horizontally inclined surfaces might also need a substantial amount of mass balancing to prevent droop and/or employ spring balancing, which may affect the dynamic behaviour. A more sophisticated solution is to provide control surfaces with hinges set back to a position just ahead of their aerodynamic and mass centres, at approximately 30%
chord. If designed carefully tabs, aerodynamic horns, mass and spring balancing
devices can all be obviated enabling a simple light weight solution with a low power
but responsive operation system to be installed. The control approximately 25% of the total tail surface area.
surface’s
area is
Another method that may be employed is to adjust the fin’s profile so that it has a
constant depth rear spar. A control surface of constant section, with hinge line parallel
to the fin’s rear spar, may then be configured. This enables the ribs to be of common construction, the spars to be of constant depth and the assembly to be a squared
arrangement (easing and standardising the joints). For an airship the impact of such deviations on the aerodynamic performance is negligible.
Assuming control surfaces with set back hinges are selected, construction may be based on a central (roughly square) torque box comprising fore and aft spars closed with outer profile thin gauged sheet skins. Leading and trailing riblets can then be
added, skinned with doped heat shrink fabric. The trailing edge will require a profiled
Structures
201
edge strip over which to wrap the fabric and the leading edge may require thin gauged sheet skins to maintain the profile. The control surfaces should be able to operate freely each way over an angle of approximately 25 degrees. Stops, however, should be incorporated at about 30 degrees to prevent runaway. The particular method of operation is largely dependent on size and role. In addition, a means to lock the control surfaces in their neutral
position should be provided for maintenance. This may also be needed when the airship is moored, to prevent unwanted motions. Conventional cable control systems have been used in the past. Whilst such methods are relatively cheap and simple, for a non-rigid airship the lost motion owing to flexibility and length may be substantial - reducing effectiveness and airship response. Also, the loads may be high and be coupled with large control movements, which place a heavy work load on the pilot, such that, in turbulent weather, the pilot's endurance may be severely reduced as a result of physical exhaustion. Modern airships are now being fitted with remote control systems that incorporate
mechanical actuators to operate the control surfaces. Such systems ensure that the degree of movement signalled by the pilot is applied at the control surface with negligible lost motion - improving control performance and airship response. It also reduces pilot work load, enabling long endurance flights with relative ease, and enables an autopilot system to be interlinked. Whichever system is adopted design of
the control surfaces and their operating method, with respect to stiffness and inertia, must avoid flutter or structural resonances. The tail surfaces may be guyed in place fairly simply using bracing cables, attached at outboard positions each side of the fin, which extend at approximately 45 degrees back to load patches on the envelope. A fail-safe system should be devised, since otherwise catastrophic failure (loss of the airship) could follow. A means to rig the cables (pretension them) is also needed. In addition, the fin should be tied to the
envelope around the root rib to enable longitudinal and transverse shear loads to be restrained. This may be done using strong cord looped several times between finger patches on the envelope and fittings on the fin and then knotted. This method may require renewal at fixed intervals in the life of the airship. Nose Structures
The bow of a non-rigid airship needs to be reinforced if:
°
The envelope’s internal pressure is not sufficient to stabilise the membrane against the aerodynamic pressure of flight, to prevent it from imploding. This is probable since otherwise the envelope’s material would have to be stronger than necessary (hence heavier), to resist the inordinately high internal pressure.
e
It is used to provide an attachment to a mooring mast.
¢
Handling/tow-in lines or a sea/ground anchor and associated winch equipment are to be fitted.
202
Airship Technology
e
It is used to install mass balance weights or ballast.
e
It is used to mount a bow thruster.
Clearly, it makes sense to provide an all embracing structure optimised for the number of influences that must be covered. In its simplest form the nose reinforcing comprises an external set of identical members arranged in a regular pattern, like the spokes of a wheel, around the envelope’s longitudinal centre-line. The members radiate from a central hub fitting, which may carry a mast attachment probe, following the envelope’s profile and extending aft far enough for the envelope’s pressure stabilised membrane to support the applied loads. Typically, this may be about 8% of the envelope’s overall length. The stiffening members may be attached to the envelope by a simple pattern of lacing, using lengths of cord. The lacing arrangement should be devised so that it can resist longitudinal loads along the stiffening members, otherwise slippage may occur, as well as tying them to the envelope’s surface. In consideration of methods for handling the airship near the ground during the mooring phase a strong bow cap may be desirable to prevent the airship from being impaled on the mast owing to uncontrollable movements. This can also be used to carry additional ballast (disposable if desired), to balance the airship, or other appendages at the bow. It may take the form of a spoked wheel, when viewed from the
front, formed to the shape of the bow, with the mast attachment probe at its centre. It can also be a simple cone, although this results in higher loads at its periphery - owing to the greater offset. To provide an aerodynamically smooth profile, it may be skinned using doped heat shrink fabric (easily repairable if necessary). Typically, the size of the bow cap may be from about 10% to 15% of the envelope’s overall diameter, designed to spread the mast reaction loads over a wide area. The bow cap should be held in position by: e
Lacing around its periphery, which primarily carry transverse shear forces - but also tie it to the envelope’s surface.
e
Longitudinal battens, which primarily carry longitudinal end loads.
The battens provide longitudinal continuity of the bow cap’s spoked members and may be laced to the envelope in similar fashion to the longitudinal stiffening members
discussed above. If the batten’s attachment to the bow cap is a pin joint (allowing the joint to articulate
as
the envelope
expands
or
contracts)
then,
in elevation,
deformations owing to transverse mast reaction loads will be as shown in Figure 7.10. As can be seen, the structure behaves in a manner similar to a four bar chain, putting compression into one side and tension on the other. At the same time the intermediate battens are displaced sideways which, owing to their shape around the envelope,
introduces torsion to those members. For these reasons the battens should be of
tubular form. It can also be seen that, on the compressive side, the lacing around the
bow cap is put into tension by the action of the battens (trying to push the cap off) and
the battens dig into the envelope at their aft end. The envelope may also buckle at this
Structures
203
point. To prevent the battens from puncturing the envelope they should be of smooth form with splayed and rounded ends. Under pure compression the envelope of a non-rigid airship may buckle at the end of its stiffening members. The load to do this is:
F = p.n.r2
(7.21)
Where: p =
envelope internal pressure.
r
envelope radius at the members aft end.
=
or" Buckle
cot oe
Bow cap
so”
S
ae i
; Ey i
Shear reaction
Mast
probe
2
_
“a
—
=< mie \
Tete
ae he
e Mast.prOves and bow CaP eas
envelope
Transverse SS applied load we
Battens pen
Envelope
Plan
Figure 7.10.
de>
View
Deformed shape of bow stiffening.
204
Airship Technology
Above this load the forward end structure merely pushes further into the envelope, like a piston in its sleeve, the load remaining essentially constant. When the load is removed the front end structure will simply pop back into shape without any apparent damage. This mechanism may be utilised in extreme conditions to cushion the mast loads, providing a means of energy absorption. Analysis of the bow reinforcing structure, if tackled using finite element techniques, should ultimately employ large displacement and non-linear theory. This is necessary in order to model its behaviour correctly on a flexible fabric structure. In addition, tests should be performed to validate those parts of the model which, owing to the unpredictable behaviour, cannot be proved by other means. This should not be underestimated, since the assets of the airship company may hinge one stormy night on the single attachment point of the bow’s stiffening structure to its mooring mast. Landing Gear
The landing gear used by early airships was a bump bag or pontoon, similar to a boat’s fender. Inflated with air this simple concept provided the necessary cushion against impacts arising from contact with the ground. Indeed, as a buoyant vehicle that floats in air, parallels for protection of the airframe and its handling in close proximity with the ground (as a ship to its pier) are appropriate. The pontoons also provided an excellent means enabling the airship to settle on water. Landing, in fact may be an inappropriate term, since airships (similar to surface ships) remain essentially afloat - never really landing - even though they may be held in close proximity with the ground. Heavier-than-air craft on the other hand convert from being air borne when in flight to being ground borne when they land; their
weight being transferred via the landing gear from the air to the ground. The landing gear more appropriately could be called a ground fender. These points are emphasised to help the reader understand essential design aspects and the nature of an airship’s undercarriage. The need for a wheeled undercarriage to reduce ground resistance arises from operational practices for take-off and landing with the airship in a state of static heaviness; requiring additional aerodynamic lift to become truly airborne. As a result, particularly if the airship does not have vectored thrust, a short rolling distance over the ground may be required, similar to that of an aeroplane, to gain sufficient air-
speed for the aerodynamic lift to become effective. The reduction of ground resistance
also helps ground handling and reduces the tendency to heel over as the airship moves sideways or weathercocks at the mast. Ground resistance may otherwise be considered useful in damping the airship’s motion across the ground.
The number and position of the units depends on overall configuration of the airframe and propeller parts needing to be protected from ground impacts and to help maintain a horizontal attitude. One unit located approximately below the airship’s
centre of gravity is the minimum requirement, providing a least weight solution. The airship remains upright owing to buoyancy, but is inclined to pitch approximately +/8 degrees and roll about 12 degrees unless other restraints are applied.
Structures
205
If the airship does not have vectored thrust capability then, if statically heavy, it either requires a rolling take-off or may be bounced off the ground using the landing
gear as a spring. In the second case a landing gear with no recoil damping may assist the take-off processes. For large airships the second method is not practical, since it requires a large ground crew. To prevent the airship from being pushed up from the ground in a bouncing mode when it lands recoil damping is necessary. The recoil rate needs to be compatible with the airship’s bouncing mode frequency, otherwise the shock absorber may still be compressed when the airship returns to the ground. Simplistically, the kinetic energy of descent to be absorbed by the landing gear is: E =(M.v2)/2
(722)
Where:
M
=_
Total Mass under consideration.
Vv
Vertical rate of descent.
For an airship with an underslung gondola supported by an external suspension system, with a wide gap between the envelope and gondola, M = mg.
Where:
mg
=
Gondola suspended mass.
For an airship with closely coupled gondola and envelope, a much greater mass must be considered: M=mg+mh+
mi
(7.231)
Where: mh
=
Hull mass (including: envelope, empennage, nose structure, etc, together with the ballonet air and lifting gas).
mi
=
Virtual inertia of the airship.
The extent to which these additional masses affect the design of the landing gear is dependent on the stiffness at the gondola’s interface with the envelope and the
stiffness of the envelope overall. The virtual inertia of an airship is a very significant term compared with heavier-than-air craft, since it can be greater than the mass of the
airship itself. This term is largely applied via the envelope. The system to be analysed may be represented as shown in Figure 7.11. y. A _A full dynamic analysis is needed to determine the maximum reaction accuratel assume the simplistic approach, however, which yields pessimistic results, is to envelope is a rigid body rigidly attached to the gondola. potential In addition to the kinetic energy the landing gear must also absorb the energy owing to maximum static heaviness:
206
Airship Technology
(7.24)
PE = (w.z)/2 Where:
w
= _ Maximum Static Heaviness.
z
=
Total landing gear vertical displacement.
The work done in compressing the landing gear is:
W = (P.z)/2 Where: P
i)
= _ Landing gear vertical reaction.
To determine the reaction we may equate the work done in compressing the landing gear with the kinetic and potential energy to be absorbed, hence:
(7.26)
(P.z)/2 = (M.v2)/2 + (w.z)/2 Therefore,
P=w+
(M.v2)/z
(7.27) Buoyancy
Envelope, nose and tail assembly mass (including contained gases)
Gravity
Figure 7.11. Dynamic simulation model.
Structures
207
The value of this reaction is very large compared with the maximum value for static heaviness. Maximum static heaviness is about 5% of the value for maximum gas lift: whereas the maximum reaction for a descent rate of, say, 5 ft/s (with the envelope treated as a rigid body rigidly connected to the envelope) is about 60% - which is the same order of magnitude as the gondola suspended weight - depending on vertical displacement. If the gondola can be isolated from the envelope the reaction would only be about 25%. Clearly the coupling between gondola and envelope is an important aspect that needs to be investigated. With a reaction this large a single landing gear unit either needs to be fitted with a large balloon tyre or more wheels, to provide a footprint large enough for the ground to accommodate the load. Since the static load is very small, and may even be zero if the airship is in a state of neutral equilibrium or statically light, brakes would be ineffective if fitted. The wheel unit should be able to castor freely through 360 degrees to any position for ground handling and take-off purposes. After take-off the wheels need to be aligned so that they trail aft, to reduce drag and be prepared for landing. This requires an additional mechanism to be devised. Compared with heavier-than-air craft the vertical displacement (tyre plus shock absorber) needs to be large to minimise the reaction load, with an element of viscous damping to increase efficiency by maintaining its peak value throughout the compressive stroke. This also enables the gas lift to play its part in arresting the envelope’s descent as the gondola’s load is transferred to the ground, the time of reaction being longer. Obviously, descent rate (being a squared term in the above equation) needs to be kept as low as possible. The authorities stipulate a minimum value of 3 ft/s. This value was used for the SKS500 airship’s original landing gear (designed in 1979 by Airship Industries), which, from personal knowledge, failed on numerous occasions including those arising from bad weather conditions at the mast. Blakemore (1927) recommends a value of 8 ft/s, resulting in loads 7 times greater. This may have been thought necessary not just for free flight landing situations but to arrest uncontrolled kiting motions at the mast. Later work done by Airship Industries showed that vectored thrust procedures for control of the final descent may enable a lesser value to
be adopted. They also found that, if the airship at its mast is kept in a state of static lightness, severe impacts are much less likely. In view of such procedures, a value of 5 ft/s may be appropriate for design. The actual value used needs to be established from trials and analysis simulating both free flight and mast mode situations. Consideration of the pitching and weather vane type motions at the mast need to be given, otherwise the landing gear may be subjected to unnecessary loads. To accommodate the airship’s motion as it circulates around the mast the wheels should be aligned athwartships and prevented from castoring. This, however, presents a problem each time contact with the ground is made. Figure 7.12a shows the airship’s motion restrained at its nose by a mast. Clearly, the wheels do not move vertically, since the airship is constrained so that it can only move in a circular are - rather they move toward and contact the ground at the angle of tangency shown in the Figure. As the tyres and shock absorber compress to absorb the energy of this motion the contact point moves forward, as shown in Figure 7. 12b, dragging the wheels at 90 degrees to
208
Airship Technology
their fixed attitude. This scrubbing action puts a heavy load across the tyres, trying to
prise them from the wheels, and, owing to trail and the locked position, introduces torque up the leg.
a) Pitching Motion at a Mast
Siz,
Ground reaction fx
Direction of motion
Rake angle -to reduce tyre
(10°)
scrubbing
Landing gear leg shock absorber unit
Castoring wheel assembly unit, aligned athwartships at the
mast to accommodate weather vane motion
Ground reaction
Tyre scrubbing due to forward movement as leg unit compresses
Figure 7.12.
Vertical landing leg attitude.
Structures
209
If the leg is raked forward so that the shock absorber aligns with the angle of tangency then the scrubbing action may be minimised. This angle, however, may also be more than the ground reaction can accommodate; depending on overall airship configuration, this angle is about 20 degrees from vertical. When frictional resistance between the tyre and the ground is low (owing, say, to ice) the ground can only react vertically. The airship when free from the mast also lands normally with vertical movement. As a result, a leg with such a large forward rake would be subjected to high bending at its upper attachment and, owing to trail and the resolved components of the load, torque up the leg opposite to that caused by the scrubbing motion. The solution ts a compromise between the forces owing to the scrubbing action and the resultant loads of the raked leg, such that the torque is neutralised for normal conditions. A rake of about 10 degrees therefore needs to be introduced.
REFERENCES Abbott, P. (1989).7he British Airship at War, 1914-1918. Lavenham: Terence Dalton Ltd. Airship Design Criteria - FAA-P-8110-2, Federal Aviation Administration, US Department of Transportation. Althoff, W.F. (1990). Sky Ships - A History of the Airship in the United States Navy.New York: Orion Books, (Crown Publishers Inc,) and Shrewsbury: Airlife Publishing Ltd. Blakemore, T.L. (1927). Pressure Airships Part 1 (Nonrigid Airships)
New
York:
The Ronald Press Company. British Civil Airworthiness Requirements - CAP471, Section Q - Non-Rigid Airships. Civil Aviation Authority, UK. Bruhn, E.F. (1973). Analysis and Design of Flight Vehicle Structures. Indianapolis: S.R. Jacobs and Associates, Inc. Burgess, C.P. (1927).Airship design. New York: The Ronald Press Company. Chamberlain, G. (1984). Airships - Cardington. Lavenham: Terence Dalton Ltd. Kinsey, G. (1988). Pulham Pigs. Lavenham: Terence Dalton Ltd. Niu, M.C.Y. (1988). Airframe Structural Design. Hong Kong: Conmilit Press Ltd. Pagon, W.W. (1927). Pressure Airships Part Il (Semirigid Airships). New York: The Ronald Press Company. Peery, D.J. and Azar, J.J. (1982). Aircraft Structures. U.S.A.: McGraw-Hill Book Co.
Williams, T.B. (1974). Airship Pilot No.28. U.K.:William Kimber and Co.
ae
7 »
ay
y >
a ‘
(
>
bee
i:
4) 4 4i
:
7
,
* aS harp esi
>
nie
-
=
P | ae)
ae
b> re >
To
=
7
; ‘
Weta
4 i@
j
a
f
;
Vek
i
eG
|
ff
S =
€te
=i
ie
{
ao
‘
:
rN
5
Ae) mers
a
AA
> vit Acta eee
aa TY
eee forme
lh po
Le
'
i
>
4
6
,
t-
ecu
he
inf
.
a
rs)
'
i. or
a veedia@ ye xioher Sie
>
‘ y
"Nom
7
tf
en,
ae
oF
7
1}
iw
ae
"a
[24
Ti)
}
“nr
_
re
ald pede
lk a el ane
maeoat /
ive
~— maAl
ie
@
2
sew Grey
yy “Se
1
‘
>
g
9
* aes rs .
——
a
af
0@
en
7
ea
i
7 |
oe
2
'
re
!
.
>"
=~ “Seq 2) rd
7
—— aD 7m
ae
‘Dat = . P42
i
a
sw
a
iw
Rai
ve
*
wit E20
itad
[
eet
routb eevd § ee
|
w
eo
Meg),
he
a
ye
ow
Cwige* Dy werd’
_—"
my
-
Md ret GAs Teh) Te BP ar. ou.” ee 5 Vanes ao i a : .
:
fe
pd tw
viv
& ia
Pao“
a
der T aT
f
eee
opi A
Aerostatics J. Craig
INTRODUCTION This Chapter provides an appraisal of airship aerostatics with an emphasis on the mass properties aspects and on operational considerations. Chapter 2 has already covered the Basic Principles of aerostatics. Chapter 13 on Performance also covers some of the same subject matter. Apologies are given now for any repetition and also if this section tends towards over-simplification. The author considers it necessary that all those involved with airships have a good understanding of the basic principles. Hence the first few paragraphs are aimed at airship novices - though they might also help clarify some misunderstandings which seem to crop up from time to time amongst those who have been involved in the industry a lot longer. This Chapter aims to give an understanding of why and how an airship can lift and to provide the means of calculating lifting performance. It also provides an awareness of many of the conditions that will affect the lift through all stages of the flight. A full list of the abbreviations and notation used is given at the end of the Chapter. The main ones though are: Lg In T M_ Ma Mg Dp pa pg pn
= = = = = = = = = =
gross lift (kg) net lift (kg) temperature (°K unless otherwise stated) total mass of gas (kg) mass of ballonet air (kg) mass of contained gas (kg) pressure (N/m’) air density (kg/m’) contained gas density (kg/m’) net or lift density (kg/m’) = pa - pg
the main subscripts used are; 0 s A
= = =
ISA, SL conditions Standard conditions (following temperature profile defined by ISA) Off-Standard conditions (same pressure but different temperature to ISA at same altitude)
212
Airship Technology
THE ATMOSPHERE An airship is dependent on the principles of buoyancy (see later) for its primary lift. The buoyancy is dependent on the density of the displaced fluid (along with the displaced volume) and, therefore, the properties of this fluid are of great concern. In the case of an airship, the fluid in which it operates is quite simply the atmosphere. The properties of the atmosphere, even in the relatively small region in which an airship operates, vary greatly. It is, therefore, extremely valuable - probably essential for the understanding of the aerostatics of an airship to be aware of the changing properties of the atmosphere. The next paragraphs will give such an understanding and explain the necessary steps to take in determining the atmospheric properties. The atmosphere is a mixture of gases, primarily oxygen and nitrogen. For the range of altitudes involved in airship operations, in fact for conventional aerodynamics generally, the composition of the atmosphere varies only slightly and may, therefore. be regarded as a homogeneous gas. For the purpose of airship analysis, only the lower region of the atmosphere - the troposphere - will be considered. This covers altitudes up to 11,000 metres. The International Standard Atmosphere (ISA)
The physical properties of the atmosphere vary greatly. For data - especially comparative data - based on the properties of the atmosphere, such as lift, it is important to have some sort of baseline or datum. This baseline is the International Standard Atmosphere (ISA). It was agreed to allow approximation of the atmospheric conditions prevailing for most of the year in temperate latitudes. The ISA is defined
by the pressure
and temperature
at mean
sea
level and by the variation
temperature with altitude. From this variation in temperature. the other characteristics of the atmosphere (pressure, p and density, p) are derived. Off-Standard atmospheric conditions can be produced by adding a temperature variation - which can be positive or negative - from the definition, the pressure at a given altitude is constant for both Standard
of
physical Data for constant ISA. By and Off.
Standard atmospheres. The data presented in this section is given for S.I. units and care must be taken to ensure that these are consistently used. ISA Mean Sea Level Conditions, (T, po, Po)
The starting point for all ideal atmospheres derived from ISA is the ISA Sea Level (SL) condition. The temperature at ISA SL, To, is 288.15°K (15°C) - note all formulae used in deriving atmospheric properties use the Kelvin scale, °K, where 0°C = 273.15°K. The pressure at SL, po, is 101325 N/m? and the density of air, pao, is
1.225 kg/m’.
AeroSstatics
213
ISA Properties (T;, ps, OP)
The density and pressure
is dependent on the temperature
at the required
altitude. ISA defines a series of temperature gradients which cover the regions of the atmosphere. temperature by 0.0065°K at a pressure
The concern here is with the lowest region, where ISA stipulates a linear gradient of -0.0065°K per metre altitude (i.e. the temperature decreases for every metre climbed). Therefore, at ISA conditions, the temperature height H, (m) is given by :
Pe=288:15:- 0.0005..17,
(8.1)
Therefore, at 1500m ISA the temperature 7, is 288.15 - 0.0065x1500 = 278.4 °K (5.25°C). The density at ISA varies exponentially with the temperature ratio, 1.e.:
; = po(T./Ts)*”
(8.2)
where the exponent 4.3 is an approximate value (see Equations 8.27-8.30). Therefore, at 1500m ISA, the density of the atmosphere is given by :
pa,= 1.225 (278.4/288.15)*? = 1.0565 kg/m°
(8.3)
The pressure at ISA similarly varies with the temperature ratio, 1.¢.
(8.4)
Ps= po(Ts/To)
where, the exponent 5.3 is also an approximate value (see Equations 8.27-8.30). Both pressure and density at any altitude in the ISA can be seen to be dependent on, and defined from, the temperature alone. Off-Standard Atmospheres
An Off-Standard atmosphere is defined by a temperature difference from ISA,
This temperature difference is constant for all altitudes. Off-Standard Properties (T4, Pay
PA
The pressure - by definition - is the same for Off-Standard and Standard = 84428 atmospheres (e.g. the pressure at ISA 1500m is the same as ISA+15 1500m
and Standard N/m’) . The density is dependent upon the ratio between Off-Standard conditions, i.e.:
214
Airship Technology bs Ts PA -o{ te)
(8.5)
where AT is the difference between Standard and Off-Standard (i.e. for ISA+15
AT =
15). For example, to derive the density at ISA-5, 1200m, the ISA conditions at 1200m
are:
T, = 288,15 —0,0065 * 1200 = 280,35° K(7,2°C) §,3
Es
= 101325
280,35)’
(ne
= 87609 N/m
r
/
(8.6)
4,3
pas
= 7225
280,35\"’ ; = 1,089 N/m (22228) /
and at ISA-5, 1200m the conditions are:
AT =-5 T, = 280,35—5 = 275,35° K(2,2°C) Pa = Ps = 87609 N/m? 280,35 280,35 = = 1,089
DEAS
ES {20.25)
ES
(8.7)
= 1,109k
F
g/m
Relative Density, o
The relative density is the ratio of the density at the required conditions, p, to the density at ISA SL, po.
o=—4
(8.8)
Values of o for a range of Off-Standard conditions have been plotted against altitude in Figure 8. 1.
Aerostatics
A B C D E
ISA + 35
F
ISA + 10
G H | J K = L M N
ISA +5 ISA
215
ISA + 30 ISA + 25
ISA + 20 ISA + 15
ISA -5 ISA - 10 ISA-15 ISA - 20 ISA - 25 ISA - 30
(km) Altitude
Figure 8.1.
Relative density, o, for a range of Off-Standard conditions against altitude.
CONTAINED GAS The gas contained within the envelope, often called the lifting gas, will also change in the same way as the atmosphere and the same calculations apply. For the purpose of this text, the contained gas is assumed to be helium. It is extremely unlikely that the gas will be 100% pure, but will contain impurities in the form of air. Due to the porosity of the envelope fabric, and also due to some leakage in the fittings, the purity will decrease during operation - and it is well to keep this in mind. To calculate the density of impure helium at ISA SL use the following formulae:
og = (kx 0.169 + (1 - k) x1.225)
(8.9)
216
Airship Technology
where 1.225 is the ISA SL density of air; 0.169 is the density of 100% pure helium at ISA SL (both figures in kg/m3); and k is percentage pure helium (e.g. if helium is 97% pure then k = 0.97). Simply calculate the density of the helium at altitude in both Standard and Off-Standard conditions by multiplying the ISA SL value by the relevant relative density. The contained gas is, however, subject to additional conditions. Firstly, it is at a higher pressure than the ambient atmosphere. This has a small effect but because it is constant, or relatively so, it can be generally ignored. Another change which cannot be ignored is superheat. This is where the contained gas is at a different temperature from the surrounding air. Not only can this have a significant effect but it can also vary throughout a single flight. Both these conditions are addressed in more detail later in this Chapter
BUOYANCY & STATIC LIFT There are several terms which are used following are brief explanations of these terms.
extensively
in aerostatics.
The
Buoyancy, B
Buoyancy is a force which is equal to the weight of fluid which a partially or completely submerged body displaces. The force acts in opposition to the weight, i.e vertically upwards. The term buoyancy is used primarily in hydrostatics, whilst in aerostatics the term static lift is almost exclusively used in its place. For consistency, static lift will be used rather than buoyancy throughout the remainder of this Chapter. Gross Static Lift, Lg
The gross static lift is the true equivalent of the term buoyancy and is equal to the weight of the air displaced by the envelope volume; Lg= Vn. pa
(8.10)
The envelope volume could be either gross ‘Vg’ (i.e. the total volume of the envelope including ballonets) or net ‘Vn’ (i.e. the envelope volume excluding ballonet volume). Which volume is used is dependent on how the envelope/gas system is viewed - i.e. as an open or closed system (explained later in this Chapter). Net Static Lift, Ln
The net static lift is the gross static lift less the weight of the contained gases within the gross envelope volume. Again this varies according to how the system is viewed:
AeroSstatics
217
In=Lg—M=Vn.pn
(8.11)
pn = pa— pg
(8.12)
where:
This net density, pr , is often called the lift density. Centre Of Buoyancy (CB)
The centre of buoyancy is the centre of gravity of the displaced fluid. It is the
point through which the static lift acts. Centre Of Gravity (CG)
The centre of gravity is the point through which the weight of an object can be taken as acting. Static Heaviness (SH)
The static heaviness is the amount by which the weight of the airship system exceeds the net static lift. The airship system comprises the airship vehicle and its systems, all the fluids (fuel, oil, water ballast etc.), the crew and payload (whether cargo, equipment or passengers) and ballast. SH
=
Washi
=
Ln
(8.13)
If the airship system weight is equal to the net static lift, then it is said to be in equilibrium (eq.). The vehicle could also be statically light (or have a negative heaviness) if the net lift exceeds the airship weight. It is generally preferred for the airship to be heavy for control purposes, but it is also usually a requirement for the airship to quickly attain, or get near to, eq. in the case of an emergency. This is normally done by dumping water ballast or in extreme cases venting the lifting gas. Airship Pressure Height
Care must be taken not to confuse airship pressure height with atmospheric pressure height, H, The airship pressure height is the height at which the ballonets are fully deflated and is therefore the maximum height to which the airship can climb without over-pressurising the envelope or venting gas.
218
Airship Technology
SUMMARY
OF AIRSHIP OPERATIONS
An airship is dependent simply on the principles of buoyancy for most of its lift. Some is supplied aerodynamically with forward speed and/or vectored thrust, but this is generally used to offset heaviness. The buoyancy is dependent upon the density of the displaced fluid, and the density of this fluid (the atmosphere) varies greatly. The following will, therefore, examine how this variation in density affects the airship, or rather how the airship copes with the changes. As already mentioned, this section will address only the basic principles and will assume primarily ISA conditions. The many other factors affecting the lifting performance will, however, be addressed later. It is explained earlier that, as an airship rises, the temperature and pressure change. Considering the contained gas, it can be seen from the Gas Law,
Dae 502) T,
T,
ee. OP MLV ee oe Pr
Ly;
ae a
(8.14) pee
that, if the temperature and pressure change, the volume will somehow have to alter. If the volume cannot change, then the pressure of the contained gas will either increase, if the temperature increases, or decrease. For most non-rigid and semi-rigid airships the differential between internal and external pressure must be kept to a fairly narrow range of values. Changes beyond these acceptable values could result in either unacceptable stresses being placed on the envelope structure (if the pressure difference is too great) or the envelope losing its structural integrity (pressure differential too low). There obviously needs to be some method for allowing this change in density, and therefore volume, without detrimentally affecting the pressure. Ballonets provide this facility. These are basically separate volumes of air within the envelope which are effectively open to atmosphere. As the density of the contained gas increases, thus decreasing the volume required, air is forced into the ballonets maintaining the
pressure differential. Similarly as the density decreases, air is pushed out of the ballonets as the volume required increases. Consider the airship at the start of a particular operation as shown in F igure 8.2. The
airship will have a certain amount of lift. This value can be calculated and measured -
to some extent approximately. The gross lift of the airship at this point is the gross
envelope volume (i.e. the volume occupied by the contained or lifting gas) times the
density of the surrounding air at that time. The net lift (i.e. the lift available for the airship, its systems and payload) is the gross lift less the weight of the contained gas, and is equal to the net volume times the net density as shown in Equation 8.15. The
net density being the difference between the air and the gas densities.
Aerostatics
219
As the airship ascends (Figure 8.2b), the contained gases will try to expand. If they cannot expand then, as can be seen from the Gas Law above, the internal pressure and therefore the pressure differential - will increase jeopardising structural integrity. In order for the volume of the contained gas to expand, air must be expelled from the ballonets. This means that the net volume, Vn, changes. As the mass of gas does not change (i.e. assuming none has leaked or been vented to atmosphere), the new volume required by the contained gas can be determined. It can be seen from the Equation 8.16 that the lift, both gross and net, does not change at any stage unless the contained gas escapes. This situation (i.e. ballonets deflating as the airship rises) continues until the airship pressure height is reached (Figure 8.2c). At this point, the contained gas has expanded to completely fill the gross envelope volume, and the ballonets are flat. Lift is still the same as at any other point during the ascent. If, however, the airship continues to rise beyond its pressure height, the contained gas cannot expand any further and so either the pressure differential increases or gas has to be vented. As already mentioned some LTA vehicles are designed for higher pressure differentials to allow for this to happen. The vast majority, though, have valves fitted to allow gas to be vented keeping the envelope pressure within limits. As the airship rises beyond its pressure height, the density of the displaced air continues to decrease. The displaced volume, however, has now reached its limit and cannot increase any further. The gross lift, Vn.pa, must decrease. In order to maintain the pressure differential within safe limits, the contained gas will be vented as the airship rises. Therefore the mass of contained gas decreases. As the lift has now decreased, the airship in effect becomes heavier. As the airship
then descends, the volume required to contain this smaller amount of gas is obviously less than during the ascent and a greater amount of ballonet fill will be required at each equivalent pressure height. Therefore, one concern of exceeding the airship pressure height is that there will be enough ballonet volume left at ground level, 1.e. the ballonets will not become 100% full before the airship reaches ground. The foregoing discussion is based on the airship operating in ISA conditions.
220
Airship Technology
Gross lift (buoyancy) = displaced volume x fluid density so: Lg = Vn. pa
where
Vn=Vg-Va
The net lift is the gross lift less the weight of the contained gases
Ln=Lg-M (a) - ‘Initial’ Condition
where
M= Mg=Vn.pg
Ln = Vn. pa— Vn. pg = Vn.(pa— pg) = Vn.pn
(8.15) Gas expands. As the mass of gas remains constant, i.e. no gas is vented
Lg = Vns. pas where
M
PSs 1
and pg = Op8y. .OPay = mie. P80 Pay
STU
P80
(b) - Airship Ascending
Figure 8.2a-b.
Aerostatic parameters as airship ascends from ‘initial’ condition to pressure height.
1 =
M
LT
Ps
ek
Aerostatics
221
At pressure height Vn = Vg and net lift is still Ln = Vn. pn Giving subscript , for conditions pressure height and , for point beyond;
at
Lg, = Vg. pa, = Vg.c7. pag Lg, = Vg. pa, = Vg.0,. pag So ALg= Lg, —Lg, = Vg. pay.(o, -0))
Ln, = Lg,-M, Ln,
=
Lg, re M,
, My =Ve. pg, = Vg.07. pg ’
M,
=
Vg. pg, =
V2.7.
P&o
ALn = Vg. pay.(o, — 0,)— Vg. pgy.(o2 -—0;) ALn = Vg. png (a, - 0)
(c) - Airship Pressure Height
Lg = Vn. pas Ln = Lg— Mgs = Vn. pas — Vn. pas = Vn. png
(8.17)
Figure 8.2c.
Aerostatic parameters as airship ascends from ‘initial’ condition to pressure height.
OTHER FACTORS AFFECTING
LIFT
The preceding explanation of the operation of an airship was extremely simplistic. There are many factors affecting the lifting performance and_ its determination. These factors can be split into two groups. Firstly, there are the variables resulting from inaccuracies in measurement of parameters such as ballonet volume, gross envelope volume, internal gas temperature and pressure. The effects of these errors will be briefly discussed. The second group of variables is related to weather effects and local variations in atmosphere. Internal Pressure
The illustrations and calculations up till now have ignored the internal pressure differential, i.c. the fact that the internal pressure is higher than the ambient. Assuming the pressure differential is 2in WG, the effect on the density of the contained gas is to increase it by 0.5% at SL to just over 0.7% at 3000m. This is
effectively negligible and, although it needs to be kept in mind, there is no need to allow for it in calculations.
222
Airship Technology Superheat
Superheat is an effect where the temperature inside the envelope is higher than the ambient air temperature. This is usually due to a greenhouse type of heating effect. Superheat causes the density of the helium to decrease. This increases its
volume and thereby reduces the ballonet volume and weight of contained air. As the weight of helium is constant, the total gas weight decreases, and because the net lift is the gross lift less the total gas weight, there is an apparent increase in net lift. As the ballonet fill decreases under superheat, this apparent increase in lift is accompanied by an apparent decrease in airship pressure height. Superheat usually occurs when the airship is on the mast and in sunlight, etc. It will reduce or disappear as the airship starts flying. As it reduces the opposite effect happens, i.e. the net lift reduces and the weight of the airship, or rather its heaviness, increases. For this reason it is important that great care is taken to ensure the effects of superheat are accounted for at all stages of a flight. Negative superheat (i.e. where the helium temperature is lower than the ambient temperature) can occur but usually not to the same extent. Weather Effects
Excluding superheat, which may be considered a weather effect, the main consideration would be in rain, snow or icing conditions. On the ground any rain, snow or ice on the envelope will add to the weight of the airship vehicle, reducing its payload. As an example, on a relatively small airship of 6500 m* volume a snow covering of 1/10 of an inch over 1/7th of its surface area will weigh over 100kg! Therefore, as much as possible should be removed prior to flight. If it is known that the proposed flight will encounter any of these conditions, then an allowance should be made. Due to the large surface area and relatively low speeds, a substantial amount of water, etc. may be collected thus increasing the heaviness of the airship.
Humidity
The effect of increasing humidity is to decrease the density of the air. This results in a reduction of the displaced mass and, therefore, the gross lift. This effect from humidity is worse the higher the temperature.
CLOSED OR OPEN
Throughout this Chapter so far, the envelope is considered as an open system,
i.e the ballonets are open to atmosphere and are not considered part of the envelope volume. There is an alternative method of analysing the airship and its aerostatics, the closed system. In the closed system, the envelope is considered as a closed volume and the ballonets as separate bags within the envelope with a variable mass. Both
systems produce identical results and both have their own advantages.
Aerostatics
223
Figure 8.3 shows and compares the two systems. It can be seen that they give the
same answer when used to derive net lift (Equations 8.18 and 8.19). The formula for the closed system continues to show that the mass of gas contained in the envelope (i.e. the mass of contained gas plus the mass of the ballonet air) can be expressed as the gross envelope volume times the contained gas density plus the ballonet volume times the net density (pa - pg) (Equation 8.20). This will be useful later. For simple lift calculations it is simpler to use the open system. When looking at airship trimming, however, it is more convenient to use the closed system.
eee
CLOSED
cree
Lg = Vn.pa
Lg = Vg. pa
Ln=Lg-M
In=Le-M
M = Mg+ Ma
M = Mg but
tit
M=Vn. 2 = =
Egg a a
eee where pn = pa- pg
_
Pe
(8.18)
M = Vn. pg + Va. oe M = Vn. pg + (Vg— Vn). pa
M = Vn.(pg- pa) + Vg.pa
Ln =Vg. pa -Vn( pg - pa) — Vg. pa Ln =Vn(pa — pg)
In=Vn.pn
note M can also be expressed as ;
M = Mg + Ma
M =Vg.pg + Va. pn
Figure 8.3.
Open and closed systems compared.
224
Airship Technology
AIRSHIP BALANCE
The basics of aerostatics and its application to airships have so far been considered. It is almost equally important to be aware of the balance or trim of the airship. As for all types of aircraft, whether fixed or rotary wing, it is vital that the centre of gravity (cg) be within a prescribed range from some datum. If the cg is outside that range, it could lead to the aircraft becoming unstable or uncontrollable with predictably disastrous results. For airships, it is the relationship between the cg and centre of buoyancy (cb) which is the important parameter (knowing the position of the cg is valueless if the cb location is not known or specified). For the airship to remain statically level (aerodynamic and thrust effects are ignored here), the cg should be directly below the cb. Any horizontal offset will result in the airship adopting a pitch angle (Figure 8.4). Theoretically this angle is the angle that the line passing through both cg and cb makes with the vertical The vertical separation between cg and cb affects the handling characteristics of the airship - the greater the separation the more stable the airship becomes, though greater control inputs are required for roll and pitch manoeuvres. The vertical relationship between cg and cb also affects the acceptable horizontal separation through simple trigonometry as is shown in Figure 8.5 (i.e. for an acceptable pitch angle a the further the cg is below the cb the greater the horizontal range can be). This is important to remember when quoting cg ranges.
Figure 8.4.
Influence of horizontal offset of eg and cb.
Aerostatics
225
During the design phase much effort should be spent ensuring that the cg is as close as possible to the optimum position when considering the loaded airship. It would be, however, extremely unusual if the cg actually ended up where required through many factors. Added to the fact that the airship is likely to have to accommodate payloads of a wide variety - some of which will result in a cg position substantially different from the design typical - it becomes clear that there should be the facility to trim or change the cg - cb position. The airship should be loaded and ballast positioned to get the cg as close to the cb as possible. This is done using some sort of weight and balance schedule and loadsheet (see Chapter 9). It is extremely likely, however, that
there will still be some trimming required to get the airship level and this can be done using the ballonets.
Figure 8.5. |The vertical relationship between cg and cb.
on the cg We now can accept (hopefully) that the static trim of the airship depends
the airship cb relationship. We will now, with the help of Figure 8.6 - which considers the airship. as a closed system - look at the components of the moment unbalance of
equal ballonet fills. The figures in the left column show the airship with effectively lift, which acts gross the is Initially, this condition will be considered. The top figure
226
Airship Technology
through the cb. The remaining three figures in the column make up the total weight which is ‘balanced’ against the gross lift. The resulting moments of the weights about the cb have to be calculated. It can be clearly seen that, irrespective of ballonet fill, the cb position is fixed (compare left and right hand columns of Figure 8.6). It can, therefore, be used as a datum from which to measure the weight moments. From the second figure it can be seen that the cg of the contained gas, irrespective of ballonet fill, is coincident with the cb and so the contained gas has a zero moment about the cb. This now simplifies the procedure considerably, as it leaves only the moments resulting from the weight of the ballonet air and those of the airship vehicle and payload which can be balanced out. Now, if the airship is balanced, then the moment resulting from the ballonet air offsets the moment of the airship, i.e.: Wa.xa = —Ws. xs Wa,.xa, + Wa,.xa, = —Ws.xs
(8.21)
However if they are not equal then a mass from the front ballonet can be transferred to the rear or vice versa, i.e air can be transferred forward or aft so:
(Wa, + AW). xa, + (Wa, — AW). xa, = —Ws. xs Wa,.xa, + Wa,.xa, — AW(xa, — xa) = —Ws. xs
There are many benefits in having fore and aft ballonets both equi-sized and equispaced about the cb. If they are not, then the resulting moment could change significantly as the ballonet fill changes. If, therefore, it is assumed that they are, then:
xa, =—xa, (see Note 1)
Wa, = Wa,
so
Wa,.xa, + Wa,.(—xa,
)— AW(xa, —(—xa, )) = —-Ws.xs
(8.23)
2AW.xa, = WS.xs AW =
Ws. soe AV.pn (see Note 2) 2xa,
Note 1.
as cb is datum, one must be +ve and the other -ve.
Note 2.
pn and not ra - see Figure 8.3, Equation 8.20.
It is, therefore, possible to use the ballonets as a fine trim within realistic limits. It
must be noted that, as the airship rises, the net density decreases and, unless the differential volume AV is increased, the moment will decrease resulting in the airship
not being statically trimmed. This condition is worsened as the stage is reached where
Aerostatics
227
one ballonet is flat. Beyond that point, not only does the net density reduce but AV also reduces.
LG
XA {MG
XAA
AA XAT
AT
XAB
AB
Sra Sy,
Ww
Figure 8.6.
Moments in a Closed System.
Figure 8.6 showed the closed system and Figure 8.7 looks at the problem as an open
system. Initially, it may appear simpler because there is one less component - the
ballonet air. If left and right hand figures are compared, however, it can be seen that now the cb does not remain in a constant position for differential ballonet fills. This means the cb position has to be recalculated and the moments taken about the new
228
Airship Technology
position. Not only is this more difficult to visualise, but there is a greater chance of errors creeping in to the calculations.
Figure 8.7.
Moments in an Open System.
DERIVATION OF GAS MASS PROPERTIES Introduction
A brief description of the atmosphere used in these calculations is given earlier in this Chapter. This section outlines the derivation used to obtain accurate densities for the atmosphere in Standard and Off-Standard conditions, and also for the contained gases within the envelope. The calculations are concerned only with the first part of the atmosphere, i.e. to an altitude of 11,000m. The air is assumed to be static (with respect to the earth), to be a perfect gas and to be free from moisture.
Aerostatics
229
To keep the text and formulae as concise as possible many abbreviations are used. A full list of these is given at the end of this Chapter. Care must be taken when using these formulae to ensure the correct units and that relevant constants are used. Three suffixes are used throughout for the various conditions considered. These are : o s a
ISA SL (Sea Level) ISA Conditions (any altitude up to 11000m) Off-Standard Conditions
Standard Atmosphere
The temperatures used in the following formulae are in°K (Kelvin scale) where the ice point (0°C) is 273.15°K. To convert from °C to °K. simply add 273.15. Atmospheric properties are defined by temperature- height profiles with the general formula;
(8.24)
T=T,+A,(H-H,) where:
T,, is temperature at height H,, at the base of layer n A,, is temperature gradient in that layer In this case, with layer 0 (0-11,000m), the formula becomes:
(8.25)
T,=T)+ A,.H
Therefore, becomes:
if Ty = 288.15°K
(15°C), An = -0.0065°K/m,
and H,, = 0 meters,
Ts = 288.15 - 0.0065H
7,
(8.26)
The pressure p, at a pressure height Hy is given by : ~8o
est Si
4 AR, 0
i
where: Po
=
Aijr pressure at Standard sea level = 101325 N/mm’
go
=
Standard sea level value for gravity = 9.80665 m/s”
A
=
Temperature gradient = -0/0065°K/m
R,
=
Gas constant for air = 287.05287 N.m/kg.K
(8.27)
230
Airship Technology
the foregoing formula is derived in ESDU 68046. Combining Equations 8.24 and 8.27 we obtain:
28815-00065H,)\
ps = 101325 |————_— 213.5
(8.28)
In the Standard atmosphere the density p, at pressure height Hp is given by :
eas
Ne
—8o
Ps = Po (2)(ae) 1,
(8.29)
where: pO go A
= __ air density at Standard sea level = 1,225 kg/m3 = Standard sea level value for gravity = 9.80665 m/s” = Temperature gradient = -0/0065°K/m
R,
=
Gas constant for air = 287.05287 N.m/kg.K
the foregoing formula is derived in ESDU 68046. Therefore
De
288.15 - 0.0065H, *°°**”””
ee
288.15
(8.30)
Off-Standard Atmosphere
In an Off-Standard atmosphere the temperature profile is defined by adding a constant increment to the ISA temperature at each pressure height. Thus;
T=T,+AT+A,(H-H,)
(8.31)
where AT is the increment from ISA. Therefore, in this case with layer 0 (0-11,000m), the formula becomes:
Ts=T,+AT+A,.H
(8.32)
where 7° = 288.15°K (15°C), A” = -0.0065°K/m, and H” = 0 m. Therefore:
T, = 288.15+ AT —0.0065H
(8.33)
The pressure in the Off-Standard atmosphere is, by definition the same as in the ISA. Therefore:
Aerostatics
Ps=DPs
231
(8.34)
From the Gas Laws:
= PT,
p>.T,
ot p, : =, 2 na
8
(8.35) |
and as p; = p2 we obtain:
PyeA dicots SsToul: THEA:
(8.36)
Relative Density
The relative density is the ratio of the density of the atmosphere at any Standard or Off-Standard condition to the density at ISA sea level, ie:
o=—
for standard conditions and
Po
5
Pa
for off standard conditions
(8.37)
Po
>
Contained Gas Conditions
The contained, or lifting gas, will obviously be lighter than air. The most common gas practically used at present is helium. It is, however, unlikely to be 100%
pure and will contain some impurities in the form of air. The purity of the gas is usually given as a percentage. For example 98% pure helium will be 98% by volume of helium and 2% by volume of air. The density of the gas at ISA SL is therefore: P total =
K.P gas POS)
Ose
(8.38)
where k is the purity in percentage expressed as decimal (e.g. for 98% pure, k = 0.98). For example, if the contained gas is 98% pure helium the density is given as p = 0.98 x 0.169 + (1 - 0.98) x 1.225 = 0.19012 kg/m’, where p,), is the density of air at ISA SL = 1.225 kg/m’ and the density of 100% pure helium is 0.169 kg/m’.
The density of the gas at the required Standard or Off-Standard condition is simply obtained by multiplying the ISA SL density by the previously obtained relative density.
232
Airship Technology
The pressure of the contained pressure is taken as the sum of the atmospheric pressure plus the pressure differential between the inside and outside of the envelope, the pressure that gives the envelope its strength. Pon
Py PT,
RD
=
eas Ap
P2
(8.39)
SO
2-7,
pP
== JO)
:
P2 >
: P|
=
p,+Ap
[8
9
(T, =T,) (8.40)
Py
Ap should always have a positive value (otherwise the envelope would attempt to collapse). Therefore, the effect of this pressure differential is to increase the density of the contained gas. As the mass of the gas does not change the volume it occupies ‘Vn’ will decrease, which will result in a reduction of both gross and net lift. The amount is small. As an example, for a pressure differential of 2in WG (498 N/m’), the effect on gas density is to increase it by less than half a percent. This would, however, approximately equate to a reduction in lift of 50kg on a 10000m” airship. As this pressure differential is more or less constant throughout normal operations it does not really need to be considered - the variation in pressure differential will have an effect in the order of 0.1% and can generally be ignored. Physical determination of the airship’s lifting capacity, through measurement, weigh-off or other means will include the effect of pressure differential. The temperature difference between that of the contained gas and that of the atmosphere (superheat) can have a greater effect. It must be considered as it will be variable not only from operation to operation but could have a significant effect during a single flight. Again from the gas law: 1g
1)
ce aps ee Pr-T; 2-72
Ree
Pi
T
l
ae Ee T, T) + Ty
(P2 = P1)
(8.41)
The effect is in reverse to the pressure differential, i.e. if the temperature differential is positive (unlike pressure it can be both positive and negative) the density decreases. As already mentioned, this temperature differential (the superheat) can be either positive or negative. Positive superheat (i.e. internal temperature higher than external) may arise from the aircraft being on ground in bright sunlight. The envelope will act
as a greenhouse and the internal temperature rises. Similarly, if the airship has been in a hangar overnight and is then brought into the sunlight prior to flight, the internal
temperature could remain below that of the atmosphere (i.e. negative superheat) for some time. As the airship flies, any temperature differential reduces and stabilises. If, however, the operation of the airship takes it from hot to cooler conditions and from
sunny to dull conditions or vice versa, then it should appear obvious that the superheat can change several times. It is therefore necessary to be aware of the superheat and its effects at all important stages of the flight (take off, landing, hover, etc.).
Aerostatics
233
An airship will start flying with some sort of superheat and this will usually disappear or reduce in the early stages of flight. If an airship has positive superheat at take off the effect is to decrease the density of the gas. This increases the net volume and as Lg =Vn. pa
and
In=Vn.pa—
M
(8.42)
this has the effect of increasing lift. During flight, however, the superheat will reduce thus having the opposite effect (i.e. of reducing lift and increasing heaviness). If this superheat is not accounted for initially, then the airship’s maximum heaviness limits may be exceeded at landing. Therefore, at take off, the amount of lift due to superheat needs to be determined. Conversely, if the airship has negative superheat initially, then the opposite happens. As the superheat wears off the airship gains lift or becomes lighter. The superheat should be considered at take off, landing and en-route to ensure the airship is within its design or operational limits at all time. The effect of the superheat is significant and it is, therefore, important that it be determined as accurately as possible.
TERMS AND ABBREVIATIONS USED The following are the terms and abbreviations used throughout this Chapter. Units and values are given primarily in the S.J. system with British/American equivalents provided in parenthesis,
A Temperature Gradient [-0.0065°K/m (-0.0019812°K/ft)] Zo _ Standard sea level value of acceleration due to gravity [9.80665 m/sec” H i. k pa Pao pas; pa, pg pg)
pgs
(32.1740 ft/sec’)] Pressure Height (m (ft)) Geopotential Height [m (ft)] Helium Purity [%] P : Atmospheric pressure [N/m’ (1b/ft”)] Atmospheric pressure at ISA Sea Level [101325 N/m’ (21 16.22 lb/ft")] Atmospheric pressure in Standard atmosphere [N/m (Ib/ft")] Atmospheric pressure in Off-Standard atmosphere [N/m* (Ib/ft”)] Pressure of contained gas [N/m? (Ib/ft*)] Gas pressure at ISA Sea Level [N/m? (Ib/ft’ )
Gas pressure in Standard atmosphere [N/m? (Ib/ft")]
Gas pressure in Off-Standard atmosphere [N/m? (Ib/ft’)] pgs Pressure differential between atmosphere & contained gas [N/m (1b/ft”)] Ap Gas constant for air [287.05287Nm/kg.K (3089.8114 Ib.ft/slug.K)] R4 Rg _ Gas constant for helium mix [Nm/kg.K (lb. ft/slug.K)] Temperature at ISA sea level [288.15 K] T, Temperature in Standard atmosphere at pressure height [K] T Temperature at pressure height in Off-Standard atmosphere [K] T,
234
Tg
T sx AT
Airship Technology Temperature Temperature and outside Temperature atmosphere
of contained gas [K] difference (superheat) between inside [K] of envelope in steady flight conditions difference of Off-Standard [K] from Standard
pa
Density of air [kg/m? (slug/ft’)]
Pao pas
Standard sea level density of air [1.225 kg/m’ (0.002376892 slug/ft*)] Density of air at pressure height in Standard atmosphere [kg/m° (slug/ft*)] Density of air at pressure height in Off-Standard atmosphere [kg/m’ (slug/ft’)] Density of contained gases [kg/m’ (slug/ft’)] Standard sea level density of contained gas [kg/m’ (slug/ft’)]
pay PS Po P§s PSA
Density of gas at pressure height in Standard atmosphere [kg/m’ (slug/ft’)] Density of gas at pressure height in Off-Standard atmosphere [kg/m’ (slug/ft’*)]
D Weight Estimates and Control J. Craig
AIRSHIP MASS PROPERTIES Introduction
This chapter aims to address most of the weights and other aspects of mass properties of airship design and operation, of which there are many. The chapter will not, however, address preliminary sizing of airships, i.e. using formulae and graphs to derive a ‘first stab’ airship weight. As explained later, the author has concerns over the validity, efficacy and accuracy of such methods. The importance of good weight estimation and control during all phases of the design of an airship, especially in the early stages, cannot be over-emphasised. Many people within the aircraft industry (and even within the LTA area) tend to accept the weight of an airship, sometimes simply due to its definition as a lighter than air vehicle, as of secondary concern. Yet the weight or amount by which the airship is overweight has a more direct effect on the performance than it would on a conventional fixed or rotary winged vehicle. Every additional kilogram of weight is one kilogram lost in available payload. Also, because an airship is very fuel efficient, any lost fuel capacity through this increased weight has an effect on endurance, which can be several orders of magnitude greater than for a heavier-than-air aircraft. Because endurance is one of the primary benefits of an airship - and in many cases its sole raison d'étre - accurate preliminary weight estimation, and strict weight control, are vital. Weight Definitions
There has been some confusion over the definitions of the various weight configurations of an airship. These should be related to the terms used for heavierthan-air aircraft, wherever possible, despite some inconsistency. The following is a list of suggested terminology for airship weights.
236
Airship Technology
Bare Weight
The ‘bare weight’ of an airship is the weight of structure, powerplant, systems and furnishings that are common to all configurations of a particular airship type. There would be no difference between an airship of a particular type destined for military use or for civil use. Bare Suspended Weight
The ‘bare suspended weight’ of an airship is the suspended weight of structure, powerplant, systems and furnishings that are common to all configurations of a
particular airship type. It should be noted that the suspended weight is a critical design criterion for airships. For this reason, a ‘suspended’ equivalent is included for most of the definitions above. Where ‘suspended weight’ is mentioned in the definitions it refers to the weight of all relevant items acting through the suspension system. Manufacturer's
Empty Weight
The ‘manufacturer’s empty weight? (MEW) is the Weight of structure, powerplant, systems, furnishings and other items of equipment that are an integral part of a particular airship configuration. Only fluids contained in closed systems are included as well. Manufacturers Suspended Empty Weight
The ‘manufacturer’s suspended empty weight’ (MSEW) is the suspended weight of structure, powerplant, systems, furnishings and other items of equipment that are an integral part of a particular airship configuration. Only fluids contained in closed systems are included as well. Basic Empty Weight
The ‘basic empty weight’ (BEW) weight of standard items.
is the manufacturer's empty weight plus
Basic Suspended Empty Weight
The ‘basic suspended empty weight’ (BSEW) is the manufacturer's suspended
empty weight plus the weight of suspended standard items.
Weight Estimates and Control
237
Standard Items
The ‘standard items’ are equipment and fluids which are not an integral part of a particular airship configuration and not a variation for the same configuration. These items include: Unusable fuel and other unusable fluids. Engine oil. Fixed ballast. Toilet fluid and chemicals. Airship emergency equipment. Galley structure, etc. Supplementary configuration).
electronic
and
avionics
equipment
(for
a
particular
Fixed mission equipment and armament not included in MEW. Operational Empty Weight The ‘operational operational items.
empty
weight’
(OEW)
Operational Suspended Empty Weight
is the basic
empty
weight
plus
(OSEW)
The ‘operational suspended empty weight’ (OSEW) empty weight plus suspended operational items.
is the basic suspended
Operational Items
The ‘operational items’ are Personnel, equipment and supplies necessary for a al particular operation but not included in the basic empty weight. These operation items may include: Disposable ballast.
Crew and baggage.
Manuals and navigation equipment. Removable service equipment for cabin, galley and bar.
Food and beverages. Usable fluids other than those in basic empty weight. Personal emergency equipment.
238
e
Airship Technology
Special mission equipment, weapon racks, etc. not included in basic empty weight. Operational Take-off Weight
The ‘operational take-off weight’ (OTW) is the operational empty weight plus disposable load at take-off. Disposable Load
The ‘disposable load’ is variable for a particular operation. This will include passengers, luggage and cargo, whether revenue or non-revenue, additional (nonessential) crew and additional carry-on mission equipment. Maximum Operational Take-off Weight The ‘maximum operational take-off weight? (MOTW) is the maximum authorised weight for take-off. It is subject to airport, operational and related restrictions.
Maximum Operational Landing Weight The ‘maximum operational landing weight? (MOLW) is the maximum authorised weight for landing. It is subject to airport, operational and related restrictions. Maximum Design Suspended Weight
The ‘maximum design suspended weight’ is the maximum weight that can be suspended or supported by the suspension system. Maximum Design Operational Weight
The ‘maximum design operational weight’ (MDOW) is the maximum weight of the airship regardless of any operational or related restrictions. It shall not be less than the MOTW. Gross Weight
The ‘Gross weight’ of an airship is the total weight of the aircraft including
enclosed gases (lifting gas and ballonet gas). It is equivalent to the gross buoyancy
plus static heaviness (or less static lightness).
Weight Estimates and Control
239
DESIGN CONSIDERATIONS FOR WEIGHT Introduction
It is not unusual, though unreasonable, to have to derive an airship weight estimate before any schemes have even been started - possibly from just a single vague paragraph. In this situation, apart from attaching as many disclaimers as possible to the weight estimate, it is advisable to be aware of the factors that will affect the weight and produce a set of assumptions that are going to be used. Pre-design Considerations
As in all vehicular projects, feasibility studies need to be carried out. For airships, widely different roles are always being proposed, many of them ambitious in nature. For all these studies, however, it is important that they are approached in the right manner. The first step is to determine the feasibility of the proposal. If the proposal is completely impractical, then it is not worth taking it any further. Many proposals, however, initially seem impractical but with some clarification can become workable. The first step must be to discuss with the client the requirements of the airship or proposed mission and agree these in writing. Only when these requirements are as clear as possible, can the criteria be derived which will allow the weight to be estimated. Design Criteria
Before any design or estimation work is carried out, it is vital that the design criteria are clearly established or derived. These criteria can be divided into those that should be considered mandatory and those that are preferred though not mandatory. The consideration of these basic design criteria should include: e
Payload (explicit and implied).
e
Endurance.
e
Performance.
¢
Operational environment (altitude and temperature ranges).
e
Maximum weights.
It is well worth the effort to ensure that these criteria are well defined and understood, as they can save a great deal of frustration at later stages of the design. Some of the above criteria are dealt with in detail elsewhere in this book but a weights overview is given below.
240
Airship Technology Payload
The payload comprises all the equipment (fixed or removable) that is required for the proposed mission plus all ancillary systems and support structure (which may be needed for the mission to be completed, even though they will not be part of the bare airship structure). The payload could be civii or military. It is just as important to account for the implied payload which is not explicitly stated but would be required to
ensure successful operations. The following lists should be considered as guidelines.
Civil Items ¢
Seating - number and type, e.g. high capacity or luxury.
e
Interior trim level.
e
Toilet facilities.
e
Additional cabin crew.
e
Galley facilities.
e
Refreshments.
e
Safety equipment - life jackets, rafts, etc.
e
Supplementary advertising - banners, night signs.
e
Additional avionics - because an item is not explicitly stated does not mean that it will not be required, e.g. night time operations on an airship only certified for
VER
will require certification with additional avionics (and lighting).
Military/Paramilitary Items
e
Supplementary electronics/mission equipment:
¢
communications, navigation, etc.
e
Electronic power requirements: the need for additional electrical generation capacity, and type of supplies required (inverters, phase sequence units, etc.).
e
Cooling requirements, both for cabin and electronics which have an effect on power requirements).
e
Additional crew.
e
Safety equipment: life jackets, rafts, etc.
e
Crew accommodation and facilities: seating, toilets, galleys, bunks, etc.
e
Underway replenishment (Unrep) requirements: dependent on endurance and operational requirements.
Weight Estimates and Control
241
e
Auxiliary fuel capacity - again dependent on operational requirements - may need several design iterations to determine the optimum.
e
Protection: both passive (stealth technology, armour, etc.) and active (missiles, etc.}.
e
In service/flight maintenance requirements.
An incomplete, or inaccurate, assessment of the payload or mission requirements could result in the weight being underestimated, which could jeopardise the feasibility of a programme. The importance of this aspect cannot be emphasised enough. Endurance
The required endurance will not only drive the fuel requirements but also payload considerations such as crew numbers. For example, will the required mission period require double crew, and facilities for crew and passengers? Or, should bunks and hot food be provided? It is important to be aware of what the operational conditions are for the endurance requirements. For example, is the endurance required for the total or the on-station requirement? If on-station, what are the transit requirements? What wind conditions are to be used? The required endurance, along with performance, will be a major deciding factor in the choice of engine configuration and type. Performance Performance, endurance and mission payload requirements can be combined to produce a mission profile. This mission profile could dictate payload and, therefore, approximate maximum lifting capability, pressure height, time on station, speed or time to station, etc. A first estimate could then be made of envelope and ballonet
volumes, gondola size, and engine specifications (numbers, power and configuration). This should be one of the first steps in deriving a new design - though it must be admitted the process is not usually this pure, and the final design includes many compromises. For a more detailed discussion please refer to Chapter 13 which is
devoted to Performance.
|
Operation Environment (Altitude and Temperature Ranges)
Careful consideration must be given to the expected range of pressure heights and temperatures including diurnal temperature ranges. This is to ensure that there is enough lift available and that the ballonet capacity is adequate to handle the required ranges.
242
Airship Technology Maximum Weights
The maximum weights should be decided early on. These should include both the maximum design operational weight and the maximum design suspended weight. Both are likely to be used early in the design phases, and it makes sense to avoid altering them at later stages. The maximum operational weight should be at least equivalent to the maximum available lift under the required operational environment. As a guide, between 0.9 and 1.0 kg/m? of envelope volume should cover most realistic cases. The maximum design suspended weight will, amongst other things, size the suspension system. Again, it is important to fix this item so as to avoid changes later which could involve a substantial amount of design and production modification whilst avoiding steps that might produce ‘over the top’ suspension systems.
WEIGHT ESTIMATION Introduction
It is not the intention of this chapter to provide readers with formulae, graphs and charts enabling them to derive, without much consideration, preliminary weight
estimates for the component parts of a modern airship. The main reason for this, as already mentioned, is validity and accuracy. It is important, far more so for an airship than a heavier-than-air aircraft, that the vehicle is not overweight. One of the primary tools in controlling the weight is an accurate and realistic preliminary estimate. Unrealistic estimates will cause trouble throughout the remainder of the design process. There are several reasons why this type of preliminary weight estimation can be
problematic. Firstly, there is a lack of recent historical data for airship weights covering non-rigid, semi-rigid and rigid airships making the derivations of parametric
links complex and sometimes tenuous. Tied into this are the technological advances and increasingly complex demands being made in and on airships. Unfortunately these requirements, whilst reducing workload and improving performance, reliability, safety and manoeuvrability, both in the air and on the ground, have produced an upward trend in the weight of modern airships. This would, when using parametric
data based on previous airships, produce a light estimate. Additionally airship design is more flexible throughout the design period with a greater number of possible configurations being considered than might be the case with a conventional fixed wing aircraft. And finally, comparisons cannot reasonably be made with other types of fixed or rotary wing vehicles.
Having stated what this section will nor be doing, it is only fair that the reader now be told what can be expected from it. The aim has been to guide the engineer or designer in deriving a weight estimate for an airship based on its role. Where the
author has considered that sufficient data are available to enable reasonable estimates
to be made, then these are given. In general, however, information is provided on how
Weight Estimates and Control
243
to derive the weight estimate, what considerations and assumptions are fair and what pitfalls to avoid. Hull Group Weight Estimation
Owing to lack of available data on modern rigid and semi-rigid hull designs this section is primarily concerned with the weight estimation of non-rigid hulls. It is hoped that this section will not only allow initial weight estimates to be derived but give the reader some guidance on the factors affecting the hull group weight. A note on scaling: for envelopes of the same profile, linear dimensions can be scaled using:
L2 = L1 (V2/V1)"3 and areas by A2 = Al (V2/V1)2. Envelope Design Criteria
Like all aspects of the airship design it is important that, in order to derive a realistic weight estimate, the criteria which are going to apply to the design of the envelope are defined. The following information should be known, or at least well documented assumptions used in its place. e
Envelope volume. For the purpose of estimating the weight of the envelope components, it would be reasonable to add 6% on to the top envelope volume. This equates approximately to just 2% on linear dimensions. This assumption should only be used for the derivation of weights and not, for example, when calculating lift.
e
Ballonet volume. This would be dependent on the range of operational conditions, primarily pressure height. Again it is recommended that a tolerance of +6% be used.
e
Ballonet configuration, e.g. fore, aft and/or pannier ballonets and their relative capacities.
e
Suspension system details. Primary maximum suspended weight (gondola). Primary suspension configuration (internal or external, adjustable, etc.). Secondary suspension/attachment requirements - will there be requirements to suspend equipment within or from the envelope or attach equipment, e.g. nightsign, to it?
© =Access and maintenance requirements. For example, if it is intended that large items are to be carried within the envelope, then possibly a large access panel would be required in the envelope fabric. Similarly, if there is a requirement for in-flight maintenance, then this implies access into and possibly onto the top of the envelope.
244
Airship Technology
Non-Rigid Envelope Weights
Table 9.1 gives a tabular weight breakdown of hull assemblies for airships with
approximate volumes of 350,000 ft’ (10,000 m*) and 2,500,000 ft* (70,800 m’). These
data are based upon well defined designs with, in the case of the smaller airship, some actual weights. The data wiil be used and expanded on in the following paragraphs. Table 9.2 contains basically the same data, but shows each major sub-assembly as a percentage of the total.
os| ie Table 9.1.
Hull Assembly Weight Breakdown
(kg)
Airship Volume >
Hull Assembly |
se
asfanlle a
110 Fadfo} ” o ys)
z °5
Q,
oC=) ga q@|Le‘s)f= S
(o)
Gaiters
> OQ Q OQ ”n2
| &o
=)No)
to
=a" pa)—e =] Comai@)i=) pe)=] Oo fa)
OQ |SsNn5 oO (o}Sc ~ ©) = aie
& Nw
os) 2
ee
927
1,244 160 ~~) ~
oe)(ame
na
= ON(a2,
Weight Estimates and Control
Table 9.2.
245
Percentage Analysis of Major Sub-Assemblies
Airship Volume >
Fabric Structure
70,800m°
WS
V2
Envelope Ballonet - 24% Ballonet - 36% Load Curtain Reinforcing Patches, Etc.
Access/Maintenance/ Gaiters
ees
Clamp Rings
24
3.0
Fabric Structure
The fabric structure is the largest component of the hull and consists of:
°
Envelope fabric.
®
Ballonet fabric.
°
Airlines.
e
Catenary curtains.
°
Patches and reinforcements, etc.
structure until it It is almost impossible to calculate a detailed weight for the ‘final’
manufactured has been fully detailed (by which time it has probably been
and
246
Airship Technology
weighed). It is hoped that the following few paragraphs will give enough information to allow a relatively accurate weight estimate in the early stages, requiring a few modifications as the information becomes more detailed. Envelope Fabric
This comprises the gores making up the envelope profile, the seams, patches and reinforcements. The envelope fabric weight is simply the surface area of the envelope multiplied by the net density of the fabric. Because of its large effect on weight, it is important that the surface area of the profile be determined as accurately as possible. If the formulae defining the profile are known, then the surface area can be obtained by integration using numerical methods such as Simpson's Rule. Great care, however, must be taken to ensure the elemental length ds and not dx is used. To use dx (which the author has seen done many times by experienced and knowledgeable engineers) would result in a significant underestimate of the surface area and, therefore, the weight. If the envelope profile has not been defined then an approximation for the surface area can be obtained from:
Area = 9.719 p (V/10.917)" where:
V =
(9.1)
Gross Envelope Volume
The net area density of the fabric, including allowances for seams, patches, etc. should be in the range 0.35 kg/m’ to 0.52 kg/m’; the greater the volume the higher the density. It is recommended that rather than use a straight line graph of density against volume that the figures in Table 9.3 be used.
Table 9.3. Area density of envelope fabric versus volume
Volume (m’)
Area Density (kg/m?)
Ballonet Fabric
The weight for ballonet fabric is more complicated to estimate accurately without details due to their relatively complex shape. Additionally there are many
Weight Estimates and Control
247
different possible ballonet configurations. The data provided here is based on ‘conventional’ forward and aft and pannier ballonets. Like the envelope fabric the ballonet fabric weight can be obtained by multiplying the surface area by a net density. The surface area for ballonets has to be considered carefully. A ballonet of a certain volume will have differing surface areas if in envelopes of different volumes (due to change in envelope profile). The formula provided here should give a pessimistic surface area. For forward and aft ballonets an approximate surface area can be obtained from:
Area = 1.07pV? and for pannier ballonets from:
Area = 3.75 p(3 358 Ba The net area density of the ballonet fabric (including seams) is between 0.275 kg/m” and 0.305 kg/m’ the larger the ballonet the lighter the net fabric density. Airlines
These are the fabric tunnels which take the air into the ballonets. represent only a small fraction of the envelope fabric weight (2%-3%).
They
Catenary This is the fabric structure from which the suspension cables are hung. The weight depends on the complexity and would vary between 9% and 14% of the envelope fabric weight, the larger the airship the higher the percentage. Patches, Reinforcement, Etc.
This will depend on the complexity of the envelope but an allowance of between 2% and 5% of the envelope fabric weight should be made. Again the larger the airship the higher the percentage. Suspension System
This covers all the items allowing the suspension of the gondola from the envelope, though excluding the catenary itself. But again, there are a variety of
possible configurations for the suspension system and for the manner in which it is attached and adjusted. Ideally, an estimate, no matter how preliminary, should be made against some sort of scheme - as with most aspects of the preliminary weight estimating discussed in this chapter - to do so without a preliminary estimate will
248
Airship Technology
almost certainly lead to difficulties. For those, however, who insist on diving into the quagmire it is suggested that a figure between 10kg and 13kg for each 1 ,000m° of envelope volume be allowed for a suspension system, the higher figure for larger and/or more complex suspension systems. These figures are based on a conventional gondola being suspended beneath the envelope and carrying all or most of the payload (making the suspended weight approximately directly proportional to the lift and, therefore, envelope volume). Nose Reinforcement Group
This is considered to consist of a nose cone and mooring probe and a structure consisting of battens or the like. There is no allowance in the figures that follow for structure, etc. for any additional systems that may be mounted on the nose. Again, basing the weight on the envelope volume, an allowance of between 17kg and 21kg per 1,000m’ of envelope volume; the higher figure should be used on smaller airships (approx. 6,000m’) whilst the lower figure should be used on larger airships (70,000m’) using a linear extrapolation for other volumes. Access/Maintenance
This will depend not only on the size of the airship but also on its intended use. For example, a small civil airship will require minimal access for ground inspection and maintenance with possibly no more than some sort of viewing aperture. A large military airship, with long endurance capabilities, would almost certainly require access not only into the envelope, but also on to the top with associated facilities allowing limited maintenance and repair. Care must, therefore, be taken when estimating the weight for this aspect of the hull. The weight allowance for the aforementioned small airship might only be 10-15kg, less than 1% of the total hull weight; whilst that for the larger airship could easily be around 300 kg, almost 3% of the total hull weight. Others
Allowance must be made for all sorts of other items fitted to the envelope, irrespective of size and complexity. Examples of such extras include fill valves, rip systems, through fabric gaiters and cable conduits. It is recommended that an allowance of 5% of the weight of the hull group obtained so far be made, and that another 3%-5% for contingency be added (inconspicuously). Paint
Painting the envelope is to be avoided. The weight effect of paint is often substantially underestimated; there can be large variations in the weight of the same area painted using the same paints on different airships. Painting cannot always be
Weight Estimates and Control
249
avoided, however, because of the requirement for environmental protection of the envelope fabric. Paint can weigh between 0.08kg/m* and 0.12 kg/m’ when on an envelope. As it will not be possible to weigh the airship after the envelope is painted, care must be taken to attempt to monitor the weight during the painting by taking samples, etc. and conducting a comparative weigh-off. Where it is intended to paint the envelope for ‘decorative’ purposes, the use of decals should be investigated. Although there may be no initial weight benefit decals can be removed when the decor changes. Paint, however, will generally be painted over, thus increasing the empty weights of the ship every time the decor is altered. Tail Group
Great care has to be taken when estimating the weight of the tail group. Owing to its large moment arm, any inaccuracy in the tail weight not only affects the weight of the airship but also has a substantial effect on the balance. If this happens, the alternatives are either to reposition the tail or the gondola so as to offset the imbalance (extremely unlikely), or to accept reduced control and performance (if imbalance is small) or add ballast. If the ballast can be mounted in the nose then the required amount of ballast will be around 80% - 100% of the tail weight error. This effectively doubles the amount by which the tail is overweight, hence the importance of realistic weight estimates (a light tail could be ballasted up to its correct weight but again this is a waste of weight that could otherwise be made available for payload). Tail Structure
The typical tail structure comprises a structural framework covered by some sort of fabric to keep the aerodynamic properties. A survey of fins over a period of two decades indicates a structural weight density in the range 4.9 - 5.4 kg/m’ (this is for the fin structure complete, i.e. including control surface). Some more recent designs, however, have a structural area weight density between 5.6 and 6.6 kg/m’. This higher weight can be put down to higher loads in the fins owing to increased performance and manoeuvrability. For a modern airship design, the weight of the fin structure (no fittings, etc. and assuming a frame and skin structure type construction) should be estimated using 4.9 kg/m° for small ‘low’ Parouanes airships, 5.9 kg/m? for smaller higher performance airships (5, 000m? -20,000m’) and 5.6 kg/m? for large airships. Although these differences may appear small their ons can be considerable. For example, an increase of 0.5 kg/m? on a 10,000m? airship could result in an increase in the total tail structure of over 70kg and moment about the CB
of 1,700 kg.m. Allowing for ballast to counter the resultant imbalance, the total
weight increase would be about 125 kg ( between | and 2 persons or enough fuel for
several hours cruising). An alternative design being promoted is the pressure contoured fin developed by an American company, ARDCO. These promise a fin structure weight up to 20% less than the figures quoted above. Additionally, if these pressurised fins are filled with
250
Airship Technology
helium, there could be an added advantage of an extra 100m? of lift volume on a
10,000m? envelope (though loss of helium from the fins would result in an imbalancing moment and must be considered carefully). Caution should be exercised
in assuming this type of design and associated weights until it has been proven in a production environment. Whichever type of design is being proposed the following components must also be considered with the fin group: ¢
Control Surfaces. Will the control surfaces be one piece or split? If split, the structure weight will be up slightly; also there will be additional control surface actuation requirements.
e
Control Surface Actuation. The actuators or actuation method is discussed later in this chapter. There must be allowances for mounting the actuators in the tail and possible lightning strike protection, especially if fins are mainly composite.
e
Strobe and Navigation Lights. Allowances must be made navigation lighting, even though it is a small amount.
e
Tail Rigging. Approximately 4 - 5.5% of the total fin and control surface structure weight should be allowed for tailfin rigging. This does not include reinforcement patches on the envelope fabric.
for strobe and
Gondola and Associated Structure Group Owing to the great flexibility available from airships, the gondola, or equivalent payload carrying module, can come in a wide variety of configurations. It is not in the
scope of this chapter to enter into any detailed investigation of the weight implications of the various configurations. What follows is a discussion of the factors that may affect the design, and therefore the weight, of the gondola structure. There is also additional information relating to ‘conventional’ suspended type gondolas. Factors to be Considered
It is important to consider carefully what is expected from the structure and
how this will affect the weight. The following is a list, which should be used as a
guide only and should not necessarily be considered comprehensive. Many proposed applications will drive the design towards a very specialised solution. e
Configuration
and Location
of Gondola.
As already mentioned,
there are
many possibilities though not explained here in detail. It is important,, however, that the static trim of the airship (when typically loaded) is considered and the
location of the gondola positioned accordingly.
Weight Estimates and Control
251
e
Suspension/Attachment to Envelope. How is the gondola to be suspended? If from suspension cables, will the cable locations and angles dictate a minimum structure length and width?
e
Purpose of Airship. \s the airship intended to carry passengers and crew, which will dictate a protected environment, or is it intended particular equipment, which could be handled by some form of structure? If an enclosed structure, then what is a realistic volume ‘cabin’ and will it require to be pressurised?
e
Location of pilots.
e
Propulsion System. Will the propulsion system be located, all or partly, in the
mission to carry skeletal for the
gondola, and if so would this result in requirements for outriggers?
e
Fuel Tankage. What volume would be required for fuel within or, in the case of external or auxiliary fuel tanks, attached to the gondola?
e
Ballast. Will space and associated structure be required for ballast, especially water?
e
Landing Gear. What is the proposed configuration of the landing gear?
¢
Other Systems. What systems will be located within the gondola structure?
e
=External mountings.
Is the proposed mission(s)
likely to require external
mounting of such items as radar and other sensors, auxiliary fuel tanks, etc.? Structural Weight
Approximation
Obviously, the weight for a gondola structure cannot be accurately estimated from a set of formulae without the inclusion of many variables. It is much more sensible to consider carefully the points raised in the preceding paragraph and sketch a scheme for the gondola that meets the derived requirements. A fair weight estimate should be obtained from this scheme and from assuming conventional aircraft structural design and technology. If advanced composite materials are proposed then the scheme should be considered more carefully and a sensible contingency be allowed.
Cabin Weights The cabin is taken as the part of the structure, which will contain the payload whether it is passengers and cargo or the mission equipment. When considering cabin structure, it is worth remembering that the larger the airship and cabin the more
complex the structure is likely to become. Great care must, therefore, be taken if
252
Airship Technology
scaling up. Weights for a wide variety of gondola sizes, from cabin volumes of less than 30m’ to greater than 350m’ have been investigated expecting some sort of square
relationship between cabin cross sectional area and weight. In fact, owing to this increase in structural complexity, the figures for all the cabin design studies had very similar weights per square metre of CSA. Allowing for flight deck structure, this came out at about 10.5 kg/m’ to 11 kg/m’. When considering the weight for the cabin structure, the following should be noted: e
Floors & Ceilings. The weight of these panels and their support can easily be underestimated.
e
Doors & Windows. The weight of these, especially doors, can be substantially greater than supposed. Account must also be taken of local reinforcing in the surrounding structure.
e
Insulation. Both noise and thermal insulation become more important as flights have longer endurance or achieve higher altitudes. Insulation could be accounted for either within the structure or in the furnishings and equipment
group. e
Bulkheads. Internal structural panels. ‘Decorative’ accounted for in the furnishings and equipment group.
bulkheads
e
Trim Panels. These are internal panels on external structure.
are
usually
e = Ballast Structure. On larger gondolas, the under-floor structure could be used for water ballast storage. If that is proposed, then allowance must be made for additional structure in this area.
¢
Equipment Mounting. If there is going to be mission equipment mounted within the cabin, then there should be suitable allowance for reinforcement. Fuel Tankage
For fuel tankage space, which will almost certainly require additional structure,
it is not unreasonable to allow between 35% and 50% of the derived linear weight for cabin structure. This figure does not allow for any fuel bladders, plumbing, etc., which are considered part of the fuel system group rather than car structure. Engine Bay
Engine bay structure must be considered very carefully as this can have a
substantial effect on weight. These considerations must include:
Weight Estimates and Control
253
e
Engine Mounting. Will it be fairly straightforward or will sub-frames. etc. be required?
e
Fire protection. Engines mounted within the gondola will need to be contained within heavy fireproof bulkheads.
e
Access space. access, etc.?
¢
Pylon. Some sort of pylon structure may be required on which to mount the propulsor. What additional equipment may be mounted on this outrigger structure, and how will it be attached to the gondola?
Will additional structure be needed for cooling, maintenance,
Other Factors
¢
Equipment Bay. Structure required to house both ship's and mission equipment. This could involve heavy avionics racking. Why do they have to use so much steel?
e
External mounting of heavy equipment such as radars, winch equipment, auxiliary fuel tanks, etc.
e
= Fitting of ships systems, i.e. electrical equipment, pneumatics systems, ECS (environmental control system), ballast tankage and equipment.
¢
Mounting of landing gear. Could involve substantial reinforcement. Propulsion System Introduction
Weight must be a serious consideration 1n the choice of powerplant and in the design of the propulsion system. It is important that the powerplant is considered, not only in conjunction with the other components of the propulsion system, but also with the fuel load for required missions. A lightweight engine may look initially attractive. If, however, its SFC in the required operating ranges is higher than a heavier engine, then the total combination of engine weight plus required fuel may be greater than might be obtained from a heavier, though thriftier, engine. This emphasises the point raised in the introduction to this chapter that it is vital that the required mission be well defined before the design should begin in earnest. Like all the sections within this the weights chapter, the choice, configuration, location, etc. of the propulsion system is considered primarily from the weights
viewpoint. The weights engineer should best propulsion system for the airship relevant designer/engineer. The weights on effects concerning weights are raised
not be making the decision as to which is the That, strangely enough, is the job of the engineer should be ensuring that the knockand, thereby, help making sure that the best
254
Airship Technology
compromise is reached in the design of the propulsion system. It is hoped that the following paragraphs will aid this aim. Powerplant Choice Considerations Type
Both piston (diesel and petrol) and turbine engines could be, and have been, used in airship applications. The following sections will outline points to be considered when choosing a type of engine. Power and Speeds
The output power of the engine as such is unlikely to concern the weights engineer unless it is grossly over the required weight In the author’s experience it is usually the reverse, i.e. the airship tends to be under-powered. This in itself, though, can lead to problems for the weights engineer as modifications to increase power output can have a substantial weight penalty, which could easily make the resultant installation heavier than if a more powerful engine had been chosen initially. Power requirements through all stages of a flight must be considered, e.g. if low cruise speeds or power requirements on station are required, it is more efficient to run engines at lower power or consider a third cruise engine. Fuel Consumption (and Fuel Type)
It is obvious that the fuel consumption is of prime importance; on an airship, however, owing to the inherent fuel efficiency (very little power requirement for lift) it is even more so than is usually granted. It is, therefore, advisable when considering propulsion systems to include an investigation of the weight of the propulsion system plus the required fuel for the typical mission - it makes little sense choosing an engine that is 50 kg lighter, for example, if it requires 100 kg more fuel for the typical
mission. Another consideration is the type of fuel. Not only is it preferable to have a less
volatile fuel on board but for airships, where it is often necessary to be able to dump ballast to equalise, it makes a great deal of sense if fuel can be used as disposable ballast. Location
Performance
is obviously
the prime criterion when
locating the airship
propulsor. The weights engineer, however, will have to think about associated effects on structure and systems when considering propulsor location.
Weight Estimates and Control
Considering the gondola mounted
255
inboard (propulsors outboard), the gondola
mounted propulsion systems have many advantages:
Minimal additionai mounting/installation structure. Near probable fuel system supply. Shorter control and indication/warning system runs. More efficient mounting hydraulic generation. Ease of maintenance required.
of ancillaries
such as electrical,
and reduction of associated
structure
pneumatic
and
and equipment
As expected it also has disadvantages:
Fire containment and protection. Adequate bulkheads and fire containment will be heavy. Noise (& Vibration Isolation). With inboard engines it will be necessary to make allowances for isolation of noise whether for civil passenger or military or para-military missions.
Transmission system. It is almost certain that at least one gearbox will be required changing direction of motion. Gearboxes can be extremely heavy (and introduce losses in engine performance).
Cooling. Inboard engines are almost certainly going to require outboard mounted cooling (& lubrication systems). Rear mounted propulsors are possible, even though they are unlikely to provide any weight advantages over outrigger mounted propulsor systems. Now we consider the Gondola, outboard
mounted; what is meant or inferred by
this is that the engine is mounted directly behind the propulsor. As the engine will be mounted in the airflow of the propeller this configuration will only be suitable for engines with a small frontal area. It has many weight advantages: Small (or no) transmission system being mounted propulsor - also improves engine efficiency.
directly
in line with
No space is required within gondola thus reducing structural mass considerably (even when allowing for increased pylon weight). Improved engine cooling thereby reducing requirements for the cooling system.
256
Airship Technology
e
Reduced cabin noise and vibration.
e
Reduced fire containment.
e
Improved access for maintenance.
Again there are disadvantages: e
Additional complexities in engine driven ancillaries such as pumps, generators, etc.
e
Generally longer control cable and fuel line runs.
Although, as can be seen in terms of weight, the advantages outweigh the disadvantages for engines which can be mounted in this way. Considering the case of envelope mounted (outriggers); although this configuration is more suited to rigid or semi-rigid airships, it could be employed on non-rigid airships. The main benefit of such an arrangement is improvement in propulsor efficiency, though other benefits include reducing the ground clearance required (reduction in landing gear weight) and improvements in noise and vibration
isolation. Disadvantages include: mounting structure required, isolation of engine mounted ancillaries, longer fuel and control cable runs, and more difficult access.
Considering the case of envelope mounted (rear of envelope); this is one of the most efficient locations for mounting a propulsor. It is also, however, one of the most awkward locations for other reasons. It is probably best suited as a location for a cruise engine. Fluids
Fluids, oil and coolants, whether trapped or disposable, should be carefully considered, especially for more powerful engine installations and for long endurance flights. Engines with a high oil consumption could require a substantial amount of lubricant to be carried on a long endurance flight. The above shows the many considerations that have to be kept in mind when designing a propulsion system. The effect of variation on weights can be great, and it is important that the weights engineers assess the implications as best, and with as much detail, as possible.
Bare engine weights can easily be obtained from trade sources or reference books.
Bare weights are limited in use as it is the installed weight which should be used for comparison. As mentioned above the combined weight of engine installation and
required mission fuel load must be considered and charts are given for this in the chapter on Performance.
Weight Estimates and Control
Table 9.4.
257
Installed engine weight allowance Small
Large
(kg/hp)
(kg/hp)
Basic Engine (Bare Engine + Mounts
0.65
0.90
Installed Engine (Basic engine + Exhaust + Cooling + Start +
0.90
1.20
|
Lube Etc.)
The figures above exclude transmission, propulsor, etc. as these depend on the configuration. This data is for reciprocating type engines. For small turbine powerplants, the basic engine weight is at the high end of the numbers, whilst the installed weight is at the lower end or even slightly below.
Propulsor
This is the propeller and duct (if relevant). Shrouding the propeller with a duct can improve propulsor efficiency whilst having additional benefits of improving safety, both directly (extremely good protection afforded to ground crew and passengers on ground) and through additional protection to the airship structure in case of propeller break up, and also reducing propeller noise. There is some weight penalty from having a duct, though any such penalty is likely to be offset by efficiency improvements.
Table 9.5.
Approximate weight of propeller and duct
Propulsor Item
Weight kg/hp
Unducted Propeller
0.14-0.21
min. 120 hp
Ducted Propeller
0.10 - 0.15
Propeller only, min. 120 hp
Duct
0.30 - 0.45
Duct only
The larger the propeller the lower the weight, whilst the reverse is true for ducts, 1.e. are the larger the duct the higher the weight. It must be emphasised that these figures available. approximate; calculated estimates must be made as soon as information is
258
Airship Technology Transmission Systems
Transmission systems can differ substantially depending on the requirements placed on them and the configuration of engine and propulsor. Where the engine is directly in line with the propeller, the gearbox, if required at all, will be a very small affair. For an inboard mounted engine and outboard propulsor, the transmission system could consist of transmission shafts, accessory drive gearbox and at least one gearbox to accommodate a change of direction. Assuming an inboard engine/outboard propulsor configuration then for approximation purposes the following can be assumed: e
Fora simple transmission system with no separate accessory gearbox, a weight of between 0.14kg/hp (for lower hp) and 0.20 kg/hp can be used.
e
For amore complex system including accessory drives, a figure of 0.25kg/hp to 0.30 kg/hp may be used. Vector Systems
Vectored thrust, where the propulsor can be swivelled, is becoming increasingly common in modern airship designs. Although it tends to be only a small part of the total propulsion system weight, between 4% and 6%, it still needs accounting for. Based on the weight of the propulsor system, the vector system weight is between 0.12 and 0.16 kg per kg of vectored mass. Fuel System
The fuel system is considered to consist of the fuel tankage (but not the structure), pipes and fittings, refuelling and fue] jettison systems, monitoring and control. Again, as the airship size or at least the fuel quantity becomes greater, the complexity of the system increases. Whereas small airships can get by with single fuels tanks with minimum transfer pipe-work etc., large systems could have many
tanks requiring complex plumbing and fuel transfer system. When estimating the tank volume, allowance must be made both for vent space and for feeder tanks, which may be required. It is also worth bearing in mind potential requirements for either auxiliary fuel tankage and refuelling during flight or making the necessary allowances for additional equipment. Control System
The control system primarily consists of the flight control system (FCS). Other control items include envelope pressurisation, trimming and engine/power control. The control systems become more complex the larger the airship and the greater the use of advanced technology. On larger airships, where the increased complexity and
Weight Estimates and Control
259
much greater cost can more easily be justified, much of the work will be done by computerised systems. Primary Flight Controls
The primary flight control systems are responsible for the control surface movement (and monitoring). Here we will consider them being moved either by cables, direct input, or by remote actuation. On smaller airships, considering the weight implications alone, a simple cable operation system has the advantage. It has, however, many more design and performance disadvantages when considering some of the more taxing operations, especially military or para-military applications. When considering the weight of a cable operation system, take into account the number of cables (redundancy in the system), pilot input (control yoke and pedal box) and emergency operation. There are more system components and requirements to be considered with a remote actuation system, namely:
e
Actuator Power: This could be either electrical or pneumatic. Currently available pneumatic actuators tend to be unsuitable for strenuous control surface applications and, therefore, an electrically powered actuator would be required. A significant weakness in having electrically powered actuators is the need for an electrical power supply. Such a supply carries a substantial weight penalty through both the power cable itself and any braiding fitted to allow for lightning strike transmittal. At least two cable runs would normally be fitted to provide the necessary level of redundancy and therefore safety in the system. A solution to this would be to provide a local (to the actuator) source of electrical power. Batteries are not only extremely heavy but are impractical owing to their limited life. A practical method would be to employ small generators, mounted in each fin, powered by pneumatics. This will obviously add the weight of these generators to the tail and a suitable pneumatic power source will be required. If these pneumatic driven generators also supplied power to the strobe and navigation lights, etc. then all the extremely heavy power cable and braiding could be replaced by extremely light plastic piping carrying the pneumatic supply. Details of possibly pneumatic power supply systems are
covered in separate sections. e
Actuators: Again, owing to the moment effect, the weight estimate for the actuators should be as accurate as possible. It is advisable to obtain an idea of the loads from the control surfaces to allow some sort of sizing of the actuators. It is also advisable to try and obtain some idea of any locally mounted electronics that may be required, such as drive units, filters, etc.
e
Signal Cables: Cabling will be required to provide control input signals to the actuators and also to return status data, such as control surface position. This can be handled by fibre optics, again relieving the need for any electrical
260
Airship Technology
contact between the gondola and the fins. Employing fibre optics will, however, require the use of some sort of computer system to translate the signals. ¢
Computer system: Some sort of computer control system will be required for an actuation system as described above. Envelope Pressurisation
The main components of the control of the envelope pressurisation are: (a) the control of the air and helium valves, (b) control of the ballonet air distribution (i.e.
directing air forward or aft, etc.) and (c) scoop operation (to gather airflow usually from behind the propulsor). All of these items require some sort of actuation and could all be controlled manually. The scoops and the tee chest (ballonet air distribution) are usually situated within, or attached to the gondola, and could therefore use electrical actuators. The air and helium valves are situated on the envelope and, as with the control surface actuation, if they were to depend on electrical actuation would require power cables - with their associated weight and lightning strike complexities. Pneumatically powered actuators are a reasonable alternative to both manual and electrical actuation. Ballast
Again the ballast dump system will require some sort of actuation. This could be manual, electrical or pneumatic. In a manual system, in addition to the cables. a weight allowance will be needed for operation handles. Both electric and pneumatic operation, while requiring actuators, can make do with simple switches. Electrical and Avionics Group Electrical
The weight of the electrical group will, like all systems, depend on what is required of it. For the purpose of this brief analysis, the electrical system is divided into: e
Generation system.
e _ Distribution system.
e
Emergency electrical equipment.
e
Miscellaneous.
Weight Estimates and Control
261
Generation System.
When considering the weight of the electrical power generation system, there are many factors that need to be taken into account. Firstly, there is the type of power required - a.c. or d.c. How is the generation system driven, by main engines alone or with an APU, etc.? How much power will be required by the ship and how much for the mission system? Also during what phases of the flight will this be required? One other factor to be considered is the fuel usage required for this power generation. In order to try and simplify preliminary weight estimates, we will relate weight to the power generation capacity (in terms of kVA). In terms of weight, the larger the generators the more efficient they become. For greater power requirements, however, it may become necessary for an APU to provide some of the power, thereby increasing installed weight. Any weights given below for the generation system are for the generators alone and any directly associated equipment (controllers, cooling, etc.). Based on generators being driven by the propulsion engines only, the weight will range from 2.3 kg per kVA - for something like 2 x 1OKVA at the lower power requirements - down to 0.875 kg at the higher (350+ kVA for example) requirements. If some of this power will be provided through an APU, then these figures could increase by over 50%-75% Distribution System
This is considered to consist of wiring, distribution hardware, CB panels, etc. For the purpose of this exercise, the wiring is split into power cables and distribution wiring. Obviously, the wiring weight depends on the location of the generations system and the systems requiring the power. It is assumed, again for
simplicity, that all the generation and most of the distribution are located in the gondola. Relating the power wiring to the power requirements, the cable weight should be between 1.2kg and 1.6kg per kVA; again the higher the power the more efficient the system becomes (i.e. 1.2kg would be the figure to use for higher electrical powers). For distribution wiring, assuming no special requirements, a figure of 5.5kg per metre length of gondola cabin can be used (length includes flight-deck and equipment bays but not propulsion area). This obviously will increase substantially if there is much electronics to be supplied. The weight of the remainder ofthe distribution system, CB panels, etc, is more risky to estimate in such a manner, but an allowance between 90kg for small airships and 350 kg for large airships should be made. The above weights relate to a ‘complex’ airship design. Small, less complex designs, such as may be required for advertising, etc. would be expected to be considerably lighter.
262
Airship Technology Emergency Electrical Equipment.
The electrical emergency equipment will depend not only on requirements but also on the configuration and degree of redundancy in the electrical system design. Emergency power requirements will include ballonet fans (to ensure envelope pressurisation is not lost), flight control system, electrical lighting, etc. Certifying authorities attitudes must also be taken into account - whether they insist on total generation failure or double engine failure, etc. For example with generation supplied from main engines and an APU total loss of power is extremely unlikely and therefore additional electrical power supplies may not be required. If, however, generation is only supplied by main engines then total engine failure and therefore loss of electrical power will have to be considered. In this case an additional power supply such as batteries will have to be provided. The sizing of the batteries will depend on requirements such as powering ballonet fans from a height. As batteries are extremely heavy, their use should be kept to a minimum. Miscellaneous Electrical Equipment
Miscellaneous electrical equipment would include lighting, power supplies for ancillary equipment (such as galley equipment) etc. This is hard to estimate without knowing the exact requirements. As it tends to be only a small part of the electrical system weight, however, it is not worth spending a great deal of effort estimating the weight. Avionics
This section is intended for ship's avionics only and not for any additional mission specific equipment. Because much of the ship’s avionics will be standard aircraft items, it is recommended that actual weights are used where possible. These can easily be obtained, either from the reference books available such as Jane’s or from the manufacturers direct. The only area where caution may be needed is the location of aerials and radar heads.
Ballast Group
Ballast is a very important component of airship design and operation - almost a necessary evil. Ballast Requirements
There are three possible reasons for carrying ballast on an airship. The first reason is the potential requirement for ballast to offset design balance deficienc ies.
Assuming an airship is designed with the gondola positioned by estimated weights to give a certain CG position relative to the centre of buoyancy (CB). If the tail structure,
Weight Estimates and Control
263
when manufactured, is overweight by 50kg and the tail is located 30m aft of the CB
then there would be an unwanted moment about the CB of 1,500kg.m equivalent to a
rearward CG shift of 0.17m at an operational weight of 8,800kg. If action was not
taken, then the ship, at normal operation conditions, would be substantially out of trim and, therefore, would be limited in its performance or capabilities. One alternative is to reposition the gondola. This, however, would be almost certainly too late as by this
stage most components would have been manufactured or procured. The only practical course of action would be to add a suitable amount of ballast at a practical location; obviously the greater the moment arm the smaller the quantity of ballast would be required. In the above example, if the ballast could be located on the nose (which was 27 m forward of the CB) then 55kg of ballast would be required. This ballast is dead weight because it reduces the amount available for payload, etc. Ballast will also be carried to keep the airship at the required heaviness range. For example, an airship with 2,500 kg of available lift but carrying only 1,800kg of payload, etc. has surplus lift of 700kg - or alternatively is statically light by 700kg. The airship can either reduce lift by venting helium, an expensive and impractical practice, or carry ballast. It is, therefore,, important that suitable stowage is provided for ballast. This stowage should be as close as possible to the payload in location to keep variations in trim conditions as small as possible. Finally, it is extremely likely that the airship will have a requirement to reduce its static heaviness quickly, especially during emergencies such as loss of engine power. One possible method is to dump fuel, but this is not always either acceptable or practical. Where fuel dump cannot be used, then a suitable amount of jettisonable ballast should be carried. For ballast, which may be jettisoned relatively regularly, it is preferable that the ballast is water. In cases of emergency, however, it may be acceptable for shot to be used. If ballast is already being carried to trim out design deficiencies, see above, it makes sense for this ballast to be disposable. Whatever form any disposable ballast takes, weights for the following items - as well as for the ballast itself -must be allowed for:
e
Storage. Whether water tanks or space for shot storage, some sort of structural allowance must be made. In addition there may be requirements for sensors to measure and display content, possible heaters in the case of water and for larger airships with multiple ballast tanks some sort of method to move ballast
between tanks. e
—Jettison. The jettisoning of any ballast must not cause any risk to the airship.
e
Jettison actuation. Some sort of jettison mechanism will be required which may be actuated manually using cables or via some sort of powered actuation, e.g. pneumatic, electric, etc.
264
Airship Technology Underway Ballast Replenishment There will be requirements, on some projects, for additional ballast to be taken
on board during flights. A simple illustration is as follows. As fuel is used, the ship will become lighter. It may then become necessary to replace the weight of the usedup fuel with ballast. There are two main methods of doing this (though additional water could also be collected relatively simply and for very little weight from the envelope). One method is to pick up water or other ballast from the surface. This could form part of an underway replenishment (Unrep) system. An alternative method is to fit a water recovery system. This condenses the moisture suspended in the engine exhaust. Ballast pick-up will require the airship to be very close to the surface for some period whilst water recovery could be carried through many stages of the operation. Thought must be given to storage of the additional ballast, whichever method is employed. Water Recovery System
A water recovery system could be used to condense the moisture suspended in the engine exhaust. It must be emphasised that water recovery is more suited to reciprocating, rather than turbine, engines. Although it would not be impossible to have water recovery from a turbine engine, its efficiency would be low. Realistically, however, this would be impractical. Theoretically, it should be possible to retrieve lkg of condensate for every kilogram of fuel burned, making it ideal for keeping to the required ship heaviness. However, this can be complex, heavy and require a considerable amount of power. For large airships (with main engine power of over 3,000 h.p.), an allowance of 0.75 kg/h.p. should be made. Insufficient work has been carried out to date to determine an allowance accurately for smaller engine sizes. Such values are, therefore, not presented here - though it must be considered impractical for smaller airships. With reference to water recovery systems, it would be wise to consider the following additional factors:
e
Volume: Designs currently proposed for water recovery systems indicate that they would occupy a considerable amount of space. Additionally, they would
tend to have large frontal areas for cooling. ¢
e °
Power Requirements: The power requirements, and type, of the power supply need to be considered. Power supply would be required by equipment such as condensers and fans. Pipe-work. Controls.
Weight Estimates and Control
265
Pneumatic and Hydraulic Systems Pneumatics
Pneumatic power can be, and currently is being, used as an alternative power supply. Many control applications, requiring actuators, can use pneumatic or electrical power. Additionally, air can be supplied for other uses such as engine starting. Pneumatic power could either be supplied from compressors or from high pressure air storage. Stored air alone gives a finite amount of power, which limits the effectiveness of the power source. As far as weights are concerned, there are, as might be expected, both advantages and disadvantages. Pneumatic power is heavier to produce. Assuming that some sort of generation capability is required (rather than pure storage only), then compressors, dryers/filters, reservoirs, control valves, etc. will be required. Distribution, especially to the envelope, is considerably lighter and has other design and operational benefits such as low maintenance and reduced metal content.
Hydraulics Hydraulic power, like pneumatics, is another suitable power source. It is more limited, however, owing to problems resulting from the hydraulic fluid itself. Also, because of possible reactions with envelope fabrics, it is not really suitable for pumping around. Further, the weight of the hydraulic fluid could have a significant effect and must be accounted for. Hydraulic power should realistically be considered. with current technology, as being suitable for higher power applications. Other Items
These are a group of items which can all too easily be omitted or underestimated when deriving the airship operational weight. They are not strictly part of the M.E.W. and, probably, would not be considered as part ofthe payload. Standard Items
Standard items are defined as equipment and fluids that are not an integral part of a particular airship configuration. They are also not a variation for the same configuration. These items include: e
_Unusable fuel and other unusable fluids.
e
=Engine oil.
e —-Fixed Ballast.
e
Toilet fluid and chemical.
e
Airship emergency equipment.
266
Airship Technology
e
Galley structure, etc.
e
Supplementary
electronic
and
avionics
equipment
(for
a
particular
configuration).
e
Fixed mission equipment and armament not included in MEW.
The first two items are part of the airship, though not part of the MEW, and their weight must be deducted from that available for payload. The remainder are usually dependent on the mission and, therefore, also dependent on the payload. They will even usually not be considered with the payload, but the necessary allowances must be made.
Payload/Mission Equipment As already mentioned above, when considering the payload, there are many factors to be taken into account other than the explicitly stated items. Civil payloads will be primarily passenger carrying, although there will always be proposals for specific applications such as logging or bulk gas carrying - these applications are too specialised to be covered here. Passenger carrying flights will obviously require seating. The type and number of seats need to be considered. For example, a luxury seat (as might be fitted to a business jet) can easily weigh 3 times the weight of a standard high density aircraft seat. Carpets, or an alternative floor covering, will almost certainly be fitted as well as decorative wall coverings. Other facilities, such as catering and toilets, will depend on the type and endurance of the flight. Provision for hot food and drinks adds a considerable amount of weight when compared to simple storage for sandwiches, etc. The same goes for toilet facilities. While a curtain and chemi-loo might suffice for military style operations, it is unlikely to be suitable for fare paying passengers. A suitable toilet installation in this case is likely to include a vanity unit, sink, mirrors etc and will require an adequate amount of water and toilet fluid. One other consideration is the possible requirement, through aviation regulations, for cabin crew.
Requirements for an explicit military or para-military fit could add a considerable amount of weight. Whenever possible, a complete list should be obtained - if not of actual equipment, then of the requirements, e.g. how many communications channels and of what sort (VHF, UHF,
secure, etc.). Further information on military and para-
military items is given earlier in this chapter. Operational Items
A definition of items comprising the operational items can be found above. Again, these can easily be omitted when determining the lift for the available mission.
It is important to make allowances for food and water for crew and passengers, especially for the longer endurance flights.
Weight Estimates and Control
267
WEIGHT MONITORING AND CONTROL Controlling the weight of an airship is a difficult task during the design and building stages. This is in common with many small or specialised aircraft where resources are tight and planned production runs are small. Again, it should be emphasised that the weight must be tightly controlled, because any overweight can have a serious effect on the airship’s capabilities, possibly limiting its feasibility. A
5% increase in airship empty 25%! The prototype methods of cheaper heavier commercial design. This must be avoided
weight could easily lead to a reduction in payload of 20design and construction will almost certainly introduce components rather than aerospace quality items into the where possible. In fact, an airship is a good platform for
prototype components.
BUILD WEIGHT CONTROL AND ACTUAL AIRSHIP WEIGHT DETERMINATION Introduction
Whilst it is necessary to know accurately the weight and CG of a complete and inflated airship, it is practically impossible to measure it physically. The actual weight of the airship can only be obtained by weighing all the components while the airship is being built. As a check, a hangar weigh-off could be carried out in as near perfect conditions as possible. Similarly, the actual volume of the envelope, and therefore the gross lift, cannot be practically measured accurately. Build Weight Control
The build weight control is a vital operation in the life of an airship; it defines the Bare Empty Weight and is the fixed point for any modifications during the ship's life. It is not enough that the items are weighed and recorded, it is also vital to know the build standard of each item. The build weight of the airship can be divided into three groups. Firstly there is the gondola weight prior to attachment to the envelope, secondly there is the hull group weight and finally the assembly and final build weights. Gondola Weight (Conventional suspended gondola).
The gondola, when complete, will be a typical complicated aircraft mix of structure and systems. The gondola can be weighed conventionally throughout it's build and before attachment to the inflated envelope. This should be done many times. Again, it is not enough just to know what the weight of the gondola is, it is also
268
Airship Technology
necessary to know what the build standard is (i.e. what is fitted, what modifications
are included, any omissions, etc). Without these data the weight is virtually useless. The starting point for the build-up of the gondola weight will be the shell or structure. This should occur as close to completion as possible but, as far as possible, without fitting any systems or ancillary structure. The build standard should be carefully noted (i.e. the revisions of the drawings and any omissions or additions). Unless the airship is in a lengthy production run - which is unlikely - this should be done for every gondola. This is because differences between the gondolas can be substantial, especially for composite construction. Because the gondola will be mounted in some sort of support trolley or jig, data and weighing procedures must be well established and used. Hull Weight
The envelope must be weighed and, like everything else, its build standard at the time of weighing must be accurately determined and weighed. All the fitments, patches, lines, etc. must be weighed as fitted, to determine the hull weight. It is impracticable for the envelope to be weighed inflated, so it must be weighed deflated. This is acceptable for weight but is no use in deriving the CG of the envelope. The latter can only be assessed through measurement after build, at which
time an approximate check of the envelope volume can also be made. A check can be made through a hangar weigh-off and a climb to pressure height. These are discussed in slightly more detail later. Final Assembly Weights
Once the gondola is attached to the envelope, it will not be possible to weigh the vehicle again. It is, therefore, extremely important to weigh and record every item or material fitted to, or removed from, the airship. This must be undertaken with the strictest discipline.
Hangar Weigh-off
A hangar weigh-off is usually undertaken immediately after completion of build. If the weight has been monitored carefully then, with comprehensive measurement, a good assessment of the lift capabilities of the airship can be made. Even if it is not realistic to have the highest level of confidence in the derived weight,
it is still possible, with some care, to carry out a check on the weight. In order to perform the weigh-off, the following steps should be taken and data recorded or derived. Use of a properly designed load-sheet will greatly help this process.
e
Derivation of the operational weight of the airship. The basic weight should be obtained from the Weight & Balance document, to which will be added the crew, fuel, disposable ballast and any other items that are onboard the airship.
Weight Estimates and Control
269
Great care must be taken to ensure the airship is to the build standard indicated
in the Weight & Balance manual, any deviations being recorded.
¢
The airship should be brought to equilibrium by the addition or removal of ballast. The amount of ballast altered should be recorded.
e¢
The airship should be trimmed using any suitable repositioning (e.g. nose to tail) and by ballonet trimming.
¢
The helium properties (purity and, if possible, volume, internal pressure differential and superheat), local atmospheric properties (temperature, barometric pressure, relative humidity, etc.) and the ballonet fill should all be recorded.
e
The net gas lift should be determined.
e
As the airship is at equilibrium, the weight should be equal to the lift.
disposable
ballast
The main areas where error could be introduced are the envelope and ballonet volumes. The envelope volume can, to a certain extent, be determined by physical measurement. The ballonets volumes are much more difficult to determine. A method of improving the level of confidence is to carry out calibration flight. One note of caution, the envelope will almost certainly increase in volume during the first several months of operation before settling down. Envelope and Ballonet Calibration
To check the envelope volume, a flight to pressure height can be made. This should be attempted in calm and dry weather to reduce the effects from moisture, turbulence, etc. Again, as this is dependent on accurate weights, a load-sheet should be accurately filled out before the flight. At pressure height, an assessment of the static heaviness should be made. As the ballonets will now be flat, a major variable is taken out of the equation and a more accurate assessment of the lift can be made.
By levelling off at pre-determined altitudes, both during the ascent and descent, it is possible to make calibration charts for the ballonets by measuring the ballonet’s fills and calculating the lift.
WEIGHTS ASPECTS OF AIRSHIP OPERATIONS Introduction
This section is not concerned with airship lifting performance. What it aims to do is highlight areas where there is some sort of weight, or weight related, input.
270
Airship Technology
Like any aircraft, an airship should be flown, or operated, within certain limits. These limits will include weight and balance or trim criteria, and could include some or all of the following: e
Maximum Design Operating Weight: This is the maximum weight at which the airship can be flown (refer to the first section of this chapter for definitions). This will have been set early in the design process allowing stress and aerodynamic calculations to proceed.
¢
Centre of Gravity Envelope: This allows the derivation of acceptable CGs for the range of weights at which the airship can be operated. Because of the inherently wide trimming ability from the ballonets, this can be much more flexible than for a fixed wing aircraft. It has its own dangers, however, in that
the correct trim can only be maintained as long as neither ballonet is flat. As soon as the airship rises beyond that point, it will become out of trim. e
Maximum Design Suspended Weight: This, in many cases, could be more important than the Maximum Design Operating Weight, because the suspension system will be designed to this figure.
e
Maximum Suspended Weight Moment: In order to stress the suspension system, the CG range at which the suspended weights are acting is also required.
e
Maximum Static Heaviness and Lightness.
It is very important, therefore, that there is some method of ensuring that these criteria are not exceeded during operation. A load-sheet completed by the pilots before flight - using weight and balance data contained within some sort of weight handbook - should be used. This is standard aircraft practice and, although there are additional complexities for airships, it should be acceptable to aircraft pilots and operators. Weight and Balance Handbook
This document should contain weights and associated CGs for the empty airship and for all the mission fits and operational items that may be fitted to the airship. An airship is likely to have a considerable number of configurations; it is
important to be able to derive accurately and quickly the airship weight and CG for
any configuration, without resorting to the design offices. Any new equipment or
revisions
to existing equipment
should
be reflected
envelopes should also form part of this document.
in the handbook.
Any CG
Weight Estimates and Control
271
Load-sheet
The load-sheet should be completed by the pilot before flight. Its purpose is to ensure that the aircraft is loaded such that it will be within operational and design limits throughout all stages of the flight. A correctly designed load-sheet should be able to address most, if not all, of the following points.
Easy input of start data - airship Basic Empty Weight and Basic Suspended Empty Weight and associated CGs or moments. Easy addition of variables such as passengers, fuel, disposable ballast, etc. to derive airship Operational Weight and Operational Suspended Weight. Determine quickly that these derived weights are within limits. Show trim condition of the airship (with neutral ballonet fills).
Show required ballonet fills necessary to bring the airship into the required trim condition.
Allow estimation of heaviness condition at take-off, landing and worst en-route stages. Allow approximate estimate of available lift to be made for comparison with previous flights. Determine the amount of disposable ballast carried, and whether enough is being carried to bring the airship to the required heaviness condition in case of emergency.
Because of the amount of information that has to be presented, it is important that the load-sheet be kept as clear as possible, with minimum use of written and calculated data and maximum use of graphical information. Providing details of producing and using load-sheets is, however, beyond the scope of this chapter.
'
ae
pic
a
TSS,
PAa [ex
Steer
90 Ae
ek te ~
PS up
f
244
ie
ee
;
wah _
jet
f
e
a.
Vip
ettat?
AD a
|
oy
(bee?
7
ci
aj par THUy
Ue
ie
ee Se
es
oe rae
#31
(aan,
ay
aaa
See
+
=
ee titans
be
Se
aes
aw se A a
Set
ix
‘
Bar
faalireg, tos rite
a
Mes
Se
py
——.
Ges
ee
ae),per
_
wa
oe A eh
7
sabaastuapaainanianias eae qi ad Cee" -corana
al -
anal ls, cue an: ue re tem vy sandy oum
>
alk
i
Shee og ifs erty, Aha Wea epe et ,ejtimets! (edomy ts an
on
9
ay) a
¢
-
=
cane
i: ane seartipoenan diigo ph
ann
7 rT
¥: Seine, ee =O
ene
‘== a2 re prin
Steele ee
7
wn
om
Voy, 854) Se ee
.)
ai a
ns
=
y
iv A)
0
2
at-ween
:
Ser
e ) gee
:
reais
“—
on hf
eos
ye
ig
ath ao
inp
aeeaeegs tie
—
=
es
WR
y=
ew @
og
Scr
Se?
a,
ean
or
eee |
: a
ress "
jm
f q
7
:
-=
AS
—“
lal
Pa
‘
‘
he
4
“>
ra
4
an
‘
al
y
v
_
—
2
7 re 7
'
Pa
_
a7
|
:
Ls bea=
=—~
7
ie
{
:
—— =
eG
(ae
\
aed
=a
-
Le
oat
ad
es
e
é
ee ps,
.
=
=
.
») Sens
Sea
a
ates
Ae
(yoO-
20
“30°
40
“50
60°
70
-80
90
Angle of incidence (degrees)
Figure 16.10.
Projected area of the Sunship (as percent of the wetted surface area) against the angle of solar incidence (Khoury, 1986).
454
Airship Technology
It is assumed, as a first approximation, that both ‘S,’ and ‘S,’ have a 27 distribution. The diffuse sky irradiation is, therefore, assumed to fall evenly on the top half of the
Sunship's surface and the reflected component to fall evenly on the lower half. The total contribution from scattered radiation ‘E,’ would be: Bore
S54 Ay 2S, AL 120 =o
AS ain
2
(16.20)
The total energy ‘E;’ falling on the Sunship becomes:
E, = Ep t+E,=A,D+ A, (Sgt S,)/2
(16.21)
In conventional airship design this energy is reflected away by the use of a white or silver coating to avoid superheating. Efficiency of the Solar power System
The available solar energy suffers a series of reductions as it is converted into propulsive power. The total efficiency of the solar power system ‘y,’ is:
Ye
=
YaXV¥oX Ve X Vg X Vp
Ya
=
Solar cell packing area efficiency
Y~
=
Solar cell conversion efficiency
Ye
=
Electrical components efficiency
Yp
=
Propulsive efficiency
(16.22)
where:
Available Propulsive Power
The power available for propulsion ‘P’ is
P= EX
(16.23)
Solar-Powered Flight Speed
A typical cubic power-speed relation for an airship is:
Pre
GA
ee
where:
Cp
fe)
=
Drag coefficient
=
Air mass density
(16.24)
Solar Power >w
—|
Wetted surface area of the hull
es
ee
&g
070
085
090
0.98
0.90
095
0.99
40
50
60
66
Solar Cell Efficiency (%) Vpmax & Vimax are the speeds obtained from direct and total irradiation respectively. €p, a, Ge, @g are the propulsion, cell area, electrical and gear efficiencies respectively.
Figure 16.11.
| Solar powered flight speed versus cell efficiency under peak solar conditions (100 W/m’ direct, 100 W/m? sky diffused, 400 W/m? ground reflected radiation) for an airship totally covered with solar cells (Khoury, 1986),
Solar Power
457
ow nmor Change in (%) SpeedoO
Figure 16.12.
Power-speed relation of the Sunship (Khoury, 1986).
Figure 16.13 illustrates this point for hourly variations in solar-powered flight speed (for latitude 20° and for solar cell conversion efficiency of 20%) calculated for direct radiation only. The influences of orientation, season, and time of day are shown not to
be too marked between the hours 8 am and 4 pm (Khoury and Mowforth 1978).
Assumed solar cell conversion
efficiency 20% (km/h) Speed Flight Solar-Powered
Time of Day (hours)
Figure 16.13.
Effect of time of day, season and orientation of the Sunship on its solar powered flight speed at latitude 20° without power contribution from the storage units
(Khoury and Mowforth, 1978).
458
Airship Technology COMPONENTS
OF THE SOLAR POWER
SYSTEM
Solar Cells
The type of solar cell suitable for airship applications will be a stable, low cost, flexible, lightweight, and high efficiency thin film (a i
Vg
~ bayarv>
ay Oe
Ss wt
‘rc
i
[e)
Detect Range (n.mi.)
nroO
30
12 é
aS a
Figure 18.15.
4Number of Rows Row Height - in
Aitrborne surveillance radar sizing.
Design Synthesis
525
The study is used to determine sensitivities of a number of element s, their weight, and cost - all as a function of the detection range. Typical ranges of values for both weight and cost are shown in Figures 16 and 17. Once again this is for L-band, but similar plots are usually prepared for other frequency bands during prelimin ary design.
Weight Ib-
50
400
150
200
250
300
350
No. of Modules
Figure 18.16.
Antenna weight.
Cost $m -
50
100
150
200
250
300
No. of Modules
Figure 18.17.
Antenna cost.
3
400
450
526
Airship Technology
The trade studies are conducted as part of the design process and are mainly concerned with the technical aspects of the configuration. There are, however, many other trade studies performed that are not directly concerned with the design approach, but rather with the complete life cycle of the airship, the reliability and maintainability of the overall configuration, the flyaway cost, the cost of operation and support in day to day flying, and many more. The cost arena is a complex one because it is governed by whether the operation is commercial or military, the specific application that the airship would be used in, and how the particular operator would break-out the cost, i.e., the cost structure. Because there are very few airships in service, in comparison with aircraft, it has not been possible to arrive at a standard way to determine these various levels of cost. To arrive at a_ preliminary understanding, one possible option is to utilise the methodology advanced by the Air Transport Association of America in the jet aircraft era. This methodology only addresses the direct operating cost, but has certainly proved useful in the aircraft world. Its use, when applied to airships, must be modified since there are a number of discrete differences that must be accounted for. For example helium usage, numbers of crew, and spares holding are all different. The parameters that are drivers in the direct operating cost (D.O.C.) calculations are utilisation, flyaway cost, crew costs, and maintenance costs. Thus direct operating cost is of extreme importance in the selling process. Competitive pressures have caused direct operating cost to play a large role early in the design of vehicles due to the influence of design on flyaway cost. A typical carpet plot is shown in Figure 18.18, and illustrates the sensitivity of the direct operating cost to parameters of utilisation and flyaway price. The strong influence of utilisation on the D.O.C. can be seen, and represents the input of the operator. The airship price has less of a slope and thus does not have as large an impact as utilisation. The D.O.C. chart is the summary of a series of sub-calculations. The overall D.O.C. is broken out into hourly and cyclical components. The hourly component uses block hours as the parameter, where ‘block’
approximates the chock-to-chock time for an aircraft. In the case of an airship, it is off-mast to on-mast time. The cyclical component represents the flight cycle, so that one flight is one cycle. The elements that enter the calculation are grouped into: (a) airship, (b) spares and ground service equipment, (c) crew, (d) maintenance, (e) finance, (f) consumables (fuel, oil, and helium), and (g) landing and handling fees. Utilisation plays a role in several of the elements. Within (a), airship depreciation is calculated and is a function
of flyaway price, salvage value, and useful vehicle life. The same calculation is made for (b) on spares and ground service equipment, with depreciation being the output. In (c) crew cost is composed of flight (pilots and attendants) and ground (crew chief, licensed engineer, and general purpose personnel). In item (d), the maintenance
component is divided into scheduled and unscheduled actions, spares inventory, and
consumable items. The section dealing with finance (e) accounts for the interest on
any part of the purchase and the insurance of the hull. The last two sections deal with
the cost of operation, viz., (f) fuel, oil, and helium costs, and the (g) landing, handling,
Design Synthesis
527
and navigation charges. Similar cost structures have also been used by the Association of European Airlines (A.E.A.) in stating vehicle requirements. The data shown in Figure 18.18 are integrated into the design process through the production cost targets and the maintenance functions. The results of studies made in this area are incorporated via a methodology called system engineering. The system engineering activities function as the integrator of the design disciplines and include risk analysis, risk management, and risk reduction. Assessment of the reliability and maintainability (R&M) characteristics is also made through these studies as they play a major role in determining the amount of maintenance required and hence the cost of maintenance. The purpose of System Engineering is to ensure success in the program through a unified approach that completely defines all requirements on the system and establishes a vehicle configuration that will be capable of meeting those requirements. It is known as a ‘front-end’ process, because most of its tasks are performed during
the initial phase of the program. The success of the program is measured in terms of cost, schedule, and technical performance. The system engineering hierarchy assists in arriving at this condition by allocating functional requirements to defined program elements. Thus system engineering involves: (a) the transformation of an operational need into a description of system performance parameters, and a preferred system configuration using an iterative process of functional analysis, optimisation, definition, and synthesis; (b) the integration of related technical parameters to assure compatibility of all physical, functional, and program interfaces; and (c) the and maintainability, reliability, producibility, of performance, integration supportability into the total engineering effort.
1800
coun tization
rlyt
1600 1400
1200
Cost $/ Operating Direct hr- 1000
800
Figure 18.18.
Direct operating cost.
528
Airship Technology TECHNOLOGY
TRENDS
The technology level associated with airships of the vintage years is lacking in comparison with that available today. As is well-known, many of the technology developments made following World War II were not readily applied to the airship. It was not until the 1970s that any attempt was made to assess the state of technology as applied to airships, and this was done through the National Aeronautics and Space Administration (NASA) Ames Research Centre. The studies performed at that time highlighted some of the possible disciplines and areas of application. On the commercial front, the British company Airship Industries also embarked on an airship development program that utilised some of the advanced materials then available. So the trend in the 1970s was to bring the airship into the modern world through the use of available technology. The use of existing technology from other areas of application did not (and does not) address the need to perform research and development for airships in their own right. This need has continued until today, and indeed is continuing. Many of the latest designs attempt to capitalise on the new technologies - available from both the aerospace industry and allied areas - to contribute to the special needs of the airship. The application of fly-by-light flight controls, the new materials used in envelope construction, the major attempt to address the ground handling of airships by use of the bow thruster, and the development of the new lightweight radial diesel engine all illustrate the focus that is being brought to bear in updating airship capabilities. Some of the areas that are of immediate interest have been on the periphery too long and require a major advance, e.g., ground handling. With the development of large airships, primarily for military applications, a number of these areas would necessarily have to be addressed, more especially cockpit design and a completely integrated flight control system. There is need to apply the lessons learned from V/STOL aircraft such as the Harrier in the integration of flight controls given that current airships use thrust vectoring as part of the flight control system. With the research and development being applied to the use of a bow thruster for masting/demasting the need becomes even more apparent. As was stated above technology development as applied to airships was close to non-existent from the 1930s until the 1970s. The only work carried out being that by
Goodyear. The NASA funded studies in the 1970s attempted to provide a comparative assessment as to the differences between the state-of-the-art in the 1930s and that prevailing in the 1970s. Two examples are taken from the NASA study to illustrate the approach taken. The first example of this difference is shown in Figure 18.19 where a typical empty weight versus volume trend is plotted for a non-rigid type airship. The trend lines represent the change in technology by comparing the trend in the 1930s with that of
the mid-1970s. The line in the centre is a mean of several non-rigid and rigid airships from the 1930s with cruise speeds ranging from 30 to 70 knots.
Design Synthesis
Oper Weight Empty vs Volume
4-
100 KTAS
Oo — 1930 SOTA ¢-
50 KTAS
Ib OWE
Volume - cu ft
Figure 18.19.
Weight trend.
fe/Swet
4-0-0001 0-0-:002 0-0:003
@-0:004
#-0:005
Area Plate Flat Equivalent fe ft sq -
1
10
Og
10°
10°
Wetted Area - Swet - sq ft
Figure 18.20.
Drag trend.
10°
10°
529
530
Airship Technology
The weight trend is typical for the 1970s for cruise speeds of 50 and 100 knots. In the lower 1970s case (50 knots cruise speed), it can be seen that the weight reduction
due to advanced technology can be expressed either in terms of a straight-forward
weight reduction (for a given volume) or a volume reduction (for a given empty weight). The weight reduction is in the order of 20 to 30 %, and is primarily attributable to the improvements made in envelope materials. The upper line (100 knots cruise speed) actually indicates a weight increase. The increase, of approximately 30%, being due to the higher aerodynamic loading associated with the higher speed compared to the lower 1930s speed. The other trend comparison is that of drag as shown in Figure 18.20, where the equivalent flat plate area (f.) is plotted against wetted area (S,,). The 1930s state-ofthe art trend is shown as a solid line and represents a mean line as in the case for weight; this time including both non-rigid and rigid types. Given that a typical 1930s airship had a value of 182 x 10° cubic feet for the wetted area, the resulting f. : Sy would be on the order of 0.0033. Basing the 1970s example on an improvement in envelope materials (which also assists in providing a much smoother surface and a lower drag figure), the corresponding f. : Swe figure would approximate to 0.0030, thus indicating an 8% reduction for the same cruise speed of 50 knots. These examples are just two among many that illustrate the changes that have taken place in the world of airship design. Other trends include improved permeability of the envelope and lower specific fuel consumption of the engines.
DESIGN SYNTHESIS The design process then is an integrated function combining each of the above considerations. The nature of the interdependency of technology is illustrated in the flow diagram of Figure 18.21, which shows the process of concept development and definition to be a highly iterative one that seeks the optimum balance between the inherent priorities of each discipline. The process also recognises that candidate concepts in the several disciplines must be integrated to provide a reasonable design compromise. At the same time, the method acknowledges that constraints that may represent a reasonable objective for one concept may be unsuitable for another due to the design approach employed. For this reason, the geometry, mass properties, aerodynamics, and propulsion constraints selection as depicted in Figure 18.21, emphasise that the overall success of the design process may be measured by a figure of merit such as minimum gross takeoff weight. It is evident that the ultimate identification of reasonable performance constraints applicable to a particular concept must come in the design and analysis of the concept. The design commences with the definition of the requirements. The vehicle is required to lift a specific payload, carry it over some distance, at a selected altitude, flying at a cruise speed, and operate at certain temperature conditions. In the first instance, it is possible that only one mission will be specified. Later, however, additional missions will be added and a number of off-design conditions studied.
Design Synthesis
Define Technology Concepts
sete Identify Concepts Acceptable ?
= Revise Approach
Discard Approach Retain for Tech Evaluation and Algorithm Development
Figure 18.21.
Technology Interdependence.
531
532
Airship Technology
The next phase begins with the process of defining the airship geometry, i.e., the actual shaping of the design concept. The first estimate at a gross weight requires: (a) the hull type to be specified; (b) the dynamic lift to be estimated; and (c) the lifting gas selected. With these parameters, it is possible to arrive at an initial estimate of the hull volume. This is because the parameters provide an indication of the characteristics of the lifting gas (as the buoyant fluid) especially its specific weight, the buoyancy ratio, and the required pressure altitude. The hull geometry is defined and developed by solving the volume equation to determine the fineness ratio for a given hull type. As the hull shape is developed, attention must be given to the control curves that define smooth longitudinal fairing and transitions. The controls then become the control points in the cross-sectional views. The design is drawn, the hull cross-sections are established, and the airship perimeter and cross-sectional area are calculated. Gondola volume is determined to verify that adequate space is available for the disposable load, i.e., payload, fuel, and ballast. The physical relationship between the gondola and envelope is checked to ascertain that the suspension system is within bounds. The design is checked for balance and the tail(s) sized. Also, a check is made into any possible physical interference between the structural elements. Propulsion/airframe integration is an integral part of the conceptual process. The propulsor ducts have to be configured, propeller(s) sized, engine type selected, and engine location determined - all to a first approximation. Each of these elements is added to the configuration within the iterative design process. If any problems become apparent, the original scaling assumptions and geometry limitations are appropriately modified and the cycle of generating new physical data and verification is repeated.
With the configuration defined to this level of detail, the aerodynamics group estimates the lift and drag characteristics. Calculation of pitching moment and other stability and control parameters is made later in the process. The weights are estimated for the major elements of the configuration. Group weights for the structural system (including envelope, tail surfaces, and gondola), and the propulsion system are derived. Additional assumptions come into play here, e.g., limit design load factor and gust load factor. These group weights are then combined with the weight of fixed equipment and the disposable load (payload, initial fuel estimate, and ballast) to produce a first estimate of the mission takeoff weight.
The baseline configuration is then ‘flown’ over the mission profile in order to determine the initial performance capability. This activity is conducted in parallel with other trade studies. Analysis of the baseline will establish the parametric field to be studied. Using the data from each of these technical disciplines, the cost of each design in the parametric set is determined as a function of structural material used, its weight, engine size and type, additional equipment used, and the design requirements, i.€., maximum speeds and loadings. The costs can be exhibited in several forms such as flyaway cost, research and development budget, and life-cycle cost (over some specified period such as 10, 15, or 20 year time-frame). Performance data are generated for a series of design points in the parametric field and these constitute the basic sizing grid or carpet plot. Evaluation of the airship is made on the basis of constant performance, constant airship weight, constant airship costs, etc., depending
Design Synthesis
533
on which is the most demanding requirement. The carpet plot is simply a method of overlaying data to compare parametric variation. The usual approach is to plot two major parameters of interest against mission takeoff gross weight (or gross lift) to form the basic sizing plot. The resulting parametric field provides the design space upon which any number of constraints may be plotted. This is illustrated in Figure 18.22 which is a development of Figure 18.11. In this case, Operating Weight Empty and Cruise Speed are the two major parameters of interest and are plotted against Gross Lift as the payoff function. Superimposed on the carpet are the constraints of Hull Volume Limit and Maximum Disposable Load. These two limits, along with the Maximum Limiting Speed, indicate the Maximum Gross Lift achievable. This design point occurs at the intersection of the Maximum Limiting Speed and the Maximum Disposable Load curves as indicated. The carpet plot also functions as a summary plot of a number of the trade studies by presenting their results in the parametric field. Direct interpolation of study results is possible when presented in this format, which aids in the design decision.
Non-Rigid Configuration
2 ,a
Range = Constant
2
@.
res
Seen
m
f
& eS €
a
= 3 =
or
o
Hull volume limit - cu >
Ib 000 1Lift Gross -
Figure 18.22.
Airship sizing matrix.
534
Airship Technology
A very brief overview of an airship design process has been provided in this chapter. The approach discussed is one among many but the resulting design should not be too dissimilar to those emerging from other methods. The level of complexity that the design synthesis can embody is only dependent upon the number of parameters that have to be studied. Thus, while the points referred to here appear to be simple, all that has been shown is a top-level example.
Index
acceleration, 175 access, 243, 248, 290-2 space, 253
trunk, 291 accommodation, 288-90 actuator disc in axial motion, 108 model, 107 theory, 107-12 added mass, 40 Aerocrane, 23, 404
aerodynamic aerodynamic aerodynamic aerodynamic
assessment, 519 control, see control data, 513 drag, see drag
aerodynamic forces and moments, 25, 33,
49-51, 79, 81, 299 in mooring, 309-14 aerodynamic lift, 20 see also lift aerodynamic parameter estimation, 59-70 aerodynamic performance, 30 aerodynamic pressure, 170 aerodynamics, 13, 17-19, 25-71 AEROS-50, 423 aerostatic lift, 6, 299, 431 aerostatic lift, see lift aerostatics, 11-17, 211-34
buoyancy, 5 parameters, 220, 221 pressure airship, 13 principles, 5 air intake, 276
air supply ducting, 277 air valves, 279 Airboat, 425
airborne early warning (AEW), 490-1, 521-7 airborne surveillance, 487
radar sizing, 524 AirCruiser, 398
Airfloat HL, 435
airlines, 247
airship design, see design synthesis Airship no. 23, 308 Airship no. p6, 308 airship operations, see operations airships categorisation, 475-6, 483
category a - no lift augmentation, 476-8, 483, 494, 495, 496 category b - partial lift augmentation, 476, 478, 483, 494, 495, 496 category c - total lift augmentation, 476, 478-9, 483, 494, 495, 496 role suitability, 492-6 spectrum of potential roles, 483-92 types, 475-82 see also non-rigid airships; pressure airships; rigid airships; semirigid airships Akron, 47, 48, 51, 54, 497 ALA600, 135 altitude effects, 9, 481 altitude range, 241
amorphous silicon cells, 462 angle of tangency, 207, 209 ANR airship, 136 antenna cost, 525
antenna frequency selection rationale, 523 antenna weight, 525
anti-piracy role, 489, 495 anti-submarine warfare (ASW), 491 apparent mass, 39-42, 77 apparent moments of inertia, 77 apparent products of inertia, 77 APU, 286, 288 aramids, 146 arched load curtains, 195 area control, 488
AS 80GD, 427 atmosphere, properties of, 212-14 atmospheric pressure height, 217
536
Airship Technology
attitude perturbation, 80 automatic flight control, 102-4, 429, 430
auxiliary thrust, 294-5 avionics, 262
axial force, 77, 78
balance, 172-3, 224-8 see also weight and balance handbook ballast, 251, 284-5 jettisoning, 263 recovery, 284-5 replenishment, 264 requirements, 262-3 weight estimation, 262-4 ballonets, 13-14, 177-9, 218, 219, 226-7, 274-6, 279, 280, 294, 323, 337, 432 calibration, 269
configuration, 243 design, 280, 281 fabric area density, 247 weight, 246-7 materials, 142, 152-3, 280 surface area, 247
trunking, 277 volume, 243
British Naval Airship no.1, 303 buckling, 185 build weight control, 267 buoyancy, 5-6, 212, 216, 218, 435 control, 15, 359-71 force, 75, 76, 79 non-neutral, 102
cabin weight, 251-2 cable control systems, 201 camera role, 486, 493
carpet plot, 519 ‘carry-over’ lift, 52 castoring, 207 catenary, 247 centre of buoyancy, 75, 217, 224-6 centre of gravity, 75, 189, 217, 224-5,
270, 318 centre of pressure, 184 check valves, 277
circulation development, 124 circulation factor, 116, 117
circumferential gores, 189 civil roles, 484, 493, 494 passenger transport, 484-5 power, 484, 487, 494, 495 research, 487
lift mechanism, 6
CL160, 436 climb, 323
pressure, 8
closed system, 222-3, 225, 227
balloon
batteries, 464-5
bending moment, 180-4 Bernoulli's equation, 25, 109, 124 Biot and Savart law, 118 blade activity factors, 121 blade element theory, 112-14 Blaugas, 17 blimp, 11 blowers, 276 boarding facilities, 169 boundary layer, 27, 37, 137 control, 371-7
growth, 376-7 bow, 201 cap, 202
reinforcing structure, 204 stiffening, 203 bow thrusters, 136 Brayton cycle, 129
clutter energy profile, 522 coastguard role, 488-9, 495 coatings for envelope materials, 148 cockpit layout, 289 CODAG (Combined Diesel and Gas turbine) powerplant, 133 coherence vs. log, 69 command, control, communications, intelligence (C31), 492 compressed air storage, 368-70 compressed storage of lifting gas, 370-1 computational fluid dynamics (CFD), 60-4 contained gas, 215-16, 219, 231-3
control aerodynamic, 340, 341 characteristics, 96-102 derivatives, 106 control surfaces, 37, 250, 322 actuation, 250
Index
arrangements, 18 notation, 80
control systems cable, 201 remote, 201
537
preliminary considerations, 507-13 trade study process, 512 Diesel cycle, 129 differential pressure, 13, 15, 179 Dinosaure, 397, 398
weight estimation, 258-60 control(s), 73-106, 282, 299, 322-3, 42833 conventional shape, 385-8
direct operating cost (D.O.C.) calculations, 526-7 disc loading, 110
costs, 502
DKBA DP-800, 423 docking systems, 315-16 drag, 28-39, 171, 372, 409
antenna, 525 direct operating cost (D.O.C.) calculations, 526-7 envelope, 503-4
ground handling, 502-3 hangarage, 503 mobility, 503 operating, 481, 502 solar power, 472 cotton, 143 counter insurrection/terrorism role, 489,
495 crane lift, 485, 493
crash landings, 294
disposable lift, 6, 9, 10
bare hull, 28-39, 39 effect of additions, 33-9 bodies of revolution, 29 control cabin, 39 control car, 37 duct, 127 form (or pressure), 27 induced, 27 lenticular hull, 383 skin friction, 29, 30, 36
tail, 39 drag coefficient, 28, 30, 37, 39, 47, 48,
crash stations, 294
crew, 288-90, 481 cruise, 332
cruise speed, 326 crystalline silicon cells, 462 customs role, 487 Cyclo-Crane, 23, 405, 406, 479
D2-bis, 413, 414 dart-shaped airships, 391 DELAG, 304 deltoid shaped airships, 390 demonstration, 499-500 descent, 323, 339
design synthesis, 507-34 baseline evaluation, 517-19
baseline layout, 514-17 baseline method for trade studies, 513
basic criteria, 509
basic requirements, 510 conceptual design project, 514 conceptual process, 511 definition of requirements, 530
drag drag drag duct
bes model, 374 reduction, 137 trend, 529, 530 drag, 127
duct thrust, 127
ducted propellers, 122-9, 257, 276, 376 Dynairship, 20, 390, 397 dynamic forces, 39-46
dynamic stability analysis, 88-96 dynastats, 395-8 early warning (dew line) function, 482 Economic Exclusion Zones (EEZ), 489 economics, 500-4 electric motors, solar powered, 463
electrical block diagram, 286 electrical bonding, 173-4 electrical emergency equipment, 262 electrical power, 285, 409 distribution system, 261, 287
integrated function, 530-4
generation system, 261 electrical system, 285-8 weight estimation, 260-2
phases, 508
electronic counter-measures (ECM), 522
538
Airship Technology
electronic equipment, 290 electronic flight control systems (EFCS), T3a3TA emergency systems, 292-4 electrical equipment, 262 escape, 171-2
longitudinal, 83-4 six degrees of freedom, 78 equilibrium control, 166-7 equipment, 481 equipment bay, 253
endurance, 171, 241, 481 energy sources, 133-5
exhaust water recovery systems, 284 external mountings, 251, 253
engine bay structure, 252-3 engine drives, 351 engine mounting, 253 engine selection, 347-50 engine specific weight vs. shaft horsepower, 131 envelope, 177-89, 222-3 area relative to sphere, 387 buckling, 431
external signatures, 482
calibration, 269 costs, 503-4
design criteria, 243 joints, 188 manufacture, 186
linearised, 83-8
fabric structure, 245-7 factorised characteristic polynomial, 90, 91 factorised transfer function numerators, 90 factors of safety, 174-6 failure effects, 170 fill valves, 290 film materials for use in laminate materials, 149
fin, 36, 37, 46 distribution, 54 efficiency factors, 53
pressurisation, 260 rip system, 293
finite element (FE) methods, 176-7, 204
shape, 73, 385 stiffness, 73 volume, 243, 431
fire risk, 339 fishery protection role, 489, 495
weights, non-rigid, 244 envelope materials, 142-3, 352, 432
planform areas, 51 fire protection, 253
flat-body, 390 flight
biaxial loading, 154 coatings, 148 compliance, 188
flight control system, 46, 258-60, 288
fabric, 246 area density versus volume, 246 structure, 277
automatic, 102-4, 429, 430 electronic (EFCS), 73, 74 flight speed, solar-powered, 448-57, 454
flexural properties, 161-2
fluid dynamic calculations, 33
at incidence, 43-4 steady turning, 44-6
fracture mechanics approach, 159
fluids, 256
load bearing component, 146 permeability, 155-6 properties, 153-62
stress-strain properties, 153
fly-by-light (FBL), 18, 73, 288, 322 fly-by-wire (FBW), 18, 73, 103, 322 flying qualities, 104 flying saucer, 20 flywheels, 468-9
tear resistance, 156-61 tear strength, 156-61
form drag, 373 free ballooning, 338-9
shear properties, 154-5
technologies, 504 uniaxial and biaxial tear testing, 161 see also specific materials equations of motion, 76-82 lateral, 86-8
freight transport, 485, 493
Froude efficiency, 110 Froude number, 65 fuel, 17, 129, 133, 134, 135, 254, 318
see also specific fuels
Index
fuel bladders, 295 fuel cells, 465-7 fuel consumption, 254, 326, 327, 347
fuel fuel fuel fuel fuel
supply, 295 system, 258, 282-3 tanks, 251, 252 usage, 345-7 weight, 350
gallium arsenide cells, 462
gamma, 304, 307 gas law, 218, 219 gas lift, 352-4 gas mass properties, 228-33 gas thrusters, 318 gas transporter, 419, 420 gas turbines, 137 gas volume changes, 274 generators, 285, 286 geometry, 515, 532 Giffard airship, 476, 497
global positioning system (GPS), 429 gondola, 35, 189-98, 247
close-coupled support system, 193-4 configuration and location, 250 connections at envelope's lower surface, 195
gravitational force, 79 gross lift, 225, 519 gross Static lift, 216 ground handling, 298-9, 333-4 costs, 502-3
future systems, 315-19 nineteenth century, 302 pioneering era, 303-4 problem survey, 299 requirements, 300-2 size effect on, 314-15 techniques, 302-14 ground manoeuvring, 301, 332 ground speed, 327
HALROP, 391, 412, 415, 432 handling qualities, 104 hangar weigh-off, 268-9 hangarage, 300-1 costs, 503 heat recovery systems, 284 heave mode, 91 heave/pitch subsidence mode, 98 heaviness, 323-5, 344
heavy lift (HL), 318-19, 434, 435 helical vortex, 126
cross sections, 191 inboard mounted, 255
helicostat, 401 heliship, 401 heli-stat, 22, 23, 318, 399, 400
non-pressurised, 191
helitruck, 399, 400
outboard mounted, 255-6 overall configuration, 190 pressurised, 191
helium, 297, 330, 352, 426 liquefied storage, 371 unit lift, 6 helium dump, 347 High Altitude Long Range Observation
shear force and bending considerations, 171 structural weight approximation, 251 structure, 192
support system, 192-8 suspended mass, i190 suspension cables, 195 suspension system, 196, 25] types, 190-2 weight, 250-3, 267-8 Goodyear blimp, 486 Goodyear GZ-22, 295 GPS (global positioning system), 429 Graf Zeppelin, 17, 133 Graf Zeppelin i, 297 Graf Zeppelin ii, 11, 482
539
Platform (HALROP), 391, 412, 415, 432
high altitude microwave-powered airship, 415 high altitude unmanned airships, 415, 42933 Hindenburg, 11, 285, 297, 482, 497 Hi-Spot, 432 hoisting station, 296 hoop tension, 182
hot-air airships, 427-8 hot-air balloon flying, 321
hovering control and response, 101-2, 333, 436 hull, 177-89
540
Airship Technology
assembly weight, 244 drag generation, 372 efficiency factors, 53 rigidity, 479-80 shape, 18, 33, 137, 385-92 weight, 243-9, 268 human-powered airships, 413-14 humidity, 222 hybrid airships, 20-3, 165, 274, 319, 391- 408 hydraulics system, 265 hydrocarbon fuel, 135 hydrogen, 133-5, 297, 417, 426 unit lift, 6
lateral mode coupling, 100 lateral response to step command input to rudder, 100 lateral stability modes, 96
lateral state matrix, 92 lateral trim conditions, 86-7 leisure role, 486, 493 lenticular geometry, 20, 380-4, 388-9, 398
lift, 211, 218, 219, 221-2, 284, 392-408, 519 aerodynamic, 20 aerostatic, 6, 299, 431
carry-over lift, 52 disposable, 6, 9, 10
efficiency, 387 ice, 329-30
immigration role, 487 induced velocity estimation, 114-18 inertia effects, 76, 77, 136 inertia forces, 294
inflight servicing, 290 internal combustion engines (ICE), 129-33 internal pressure, 6, 221, 432 International Standard Atmosphere (ISA), 9, 10, 14, 212-13, 219 Joule cycle, 129 kerosene, 134
kinetic energy, 205 kinked envelope, 182-4 Kutta condition, 124 Kutta-Zhukowsky theorem, 115 La France, 409 laminate materials, 145-53
bonding and joining, 151-2 fibre properties, 146 film materials for use in, 149 gas retention component, 147-8 improved, 162 potential of, 152, 162 sailcloth as used in, 162
weathering component, 150 landing, 204, 302, 321, 325, 327, 330, 332, 338, 339-44 landing gear, 204-9, 251, 253 landing leg attitude, 208 Laplace transform, 89
performance, 269-71
lift augmentation
partial, 476, 478, 483, 494-6 total, 476, 478-9, 483, 494-6 lift/drag ratio, 134 lifting body concept, 20, 398 lifting body hybrid, 19, 20 lifting gas, 273-4, 426-8 heating, 363-8 management, 297 manipulation, 359-71
pressurisation, 360-3 lifting rotor hybrid, 22 lift-off, 301-2 lightness, 324-5, 344, 346 lightning strike, 173-4 liquefied helium storage, 371 load factors, 170 load-sheet, 270, 271 loading/unloading, 301, 318-19, 435 Long Endurance Manned Solar (L.E.M.S.) airship, 412
longitudinal gores, 188 longitudinal pendulum mode, 92 longitudinal response to elevator, 96 to thrust, 98
longitudinal stability modes, 93 longitudinal state equation, 89 longitudinal state matrix, 91 longitudinal surface length, 189 longitudinal tension, 186 longitudinal trim conditions, 84 Loral Goodyear (GZ-22), 15
Index
LORAN, 326 Los-Angeles, 307 LOTTE, 410 LOTTE-1, 411 LOTTE-2, 411 LOTTE-3, 411 low altitude unmanned airship, 428-9 low speed control, 307-9, 316-18 LTA 20-1, 407, 408 LZ-1, 303, 480 LZ-2, 303 LZ-3, 303 LZ-4, 303 LZ-8, 304 LZ-10, 304 LZ-127, 480 LZ-No 7, 424 LZ-NT, 423
MLA-32-A, 20 mobility costs, 503 moment equations, 78 monocoque ‘internal’ structure, 420 monocoque rigid construction, 419-20
Mach number, 65, 111, 113 Macon, 307, 497
neoprene, 150
Magnus Aerolift, 407 Magnus effect, 478 maintainability, 527 maintenance, 243, 248, 290-2
market analysis, 499 market potential, 496-500 marketing, 500 mass distribution, 172-3 mass properties, 235-8 mast motions, 167-9, 207, 208 materials, 141-64 textile, 141-53 see also specific materials and applications Mayfly, 304
MCP, 346, 347, 349 Megalifter, 20, 390, 397
541
mooring, 18, 301 forces and moments in, 309-14 future solution, 319
height, 305 nose attachment, 13
single-point, 305, 519 mooring mast, 304-7 multi-balloon, 392
multi-hull airships, 390, 391 multi-lobe, 398
National Physical Laboratory, 33 natural fibres, 142 navigation lighting, 250 net static lift, 216 Newton's laws, 25 Newton's second law of motion, 78
non-neutral buoyancy, 102 non-rigid airships, 11, 15, 141, 165, 218,
299, 418 structures, 177-209
Norge, 301 normal force, 78
nose mooring attachment, 13 nose reinforcement group, 248 nose rip assembly, 293 nose splines, 13 nose structures, 201-4 nuclear power, 299, 417
nylon, 143 Obelix, 402
methane, 426
Off-Standard atmospheres, 213-14, 230-1 Off-Standard properties, 213-14 oil reservoirs, 283
methane gas transporter, 426 microwave power, 415 military roles, 484, 490-2, 496
open system, 222-3, 228 operation environment, 241 operational analysis, 499
mine countermeasures (MCM), 492,
operations, 218-19 weight aspects, 269-71 oscillatory roll mode, 95 Otto cycle, 129
membrane stress, 182
metalclad airship, 11, 422
498-500 mission considerations, 345
mission equipment, 266 mission requirement, 354
542
Airship Technology
paint, weight effect, 248-9 parametric analysis, 519 paramilitary roles, 484, 488-9, 495 passenger transport role, 493
pressure difference, 6-7, 13 pressure distribution, 26, 35, 36, 38 pressure height, 7, 9, 217, 337-8
payload, 240-1, 266, 318, 345, 434-6, 481
pressure stabilised envelopes, 186 pressurisation system, 178 pressurised tube structure, 425 prime movers, 129-33
Pegasus, 388
promotional role, 486, 493
performance, 241, 325, 326, 345-57, 480-1
propane, 133 propeller blades, 112 propeller characteristics, 377 propeller loading, 110 propeller performance comparison of theory and practice,
patches, 247 Patoka, 307
personnel transfer, 295 phase angle, 69 Piasecki Heli-Stat, 318
Piasecki project, 479 piloting, 321-44 essential requirements, 322 training, 321 pitch, 121, 126, 323 pitch angle, 224-5, 323 pitch attitude response, 90 pitch-incidence motion, 99 pitch subsidence, 91 pitched flight, 380 pitching moment, 79, 305 pneumatics system, 265 police role, 487 polyamides, 146 polycrystalline cells, 463 polyester, 143, 144, 147, 152
polyethylene, 147 polyimides, 147 polyurethane, 150, 152 polyvinylfluoride (PVF), 150 potential energy, 205-6 power and shaft speed relation, 129 power off-take, 345, 350-1
power requirement using stern propulsion, 378 power source, 409-17 power storage, 464 comparison of systems, 469 powerplant choice considerations, 254-7 powerplant design, 299 pressure external, 6 internal, 6, 221, 432 pressure airship, 11, 13, 14
pressure calculation, 180 pressure coefficient, 25 pressure control, 14, 273-80
118-22 estimation, 114 propellers, 257, 351
ducted, 122-9, 257, 276, 376 with reversing/vectoring facility, 316 propulsion/airframe integration, 532 propulsion system, 107-39, 251 geometry, 81 weight, 253-8, 350
propulsive efficiency, 111 propulsive force, 107 propulsor, 107, 257
loading, 112 tail mounting, 136 prototype engineering manpower, 509 public perception, 497 purge valves, 290 pusher propulsion, 276 pylon, 253
R-25, 308 R-26, 308 R-29, 308 R-31, 298, 308 R-33, 300, 306-8 R-34, 308, 485 R-38, 308, 418, 497 R-80, 308 R-100, 309, 418, 497 R-101, 306, 307, 309, 418, 497 radar cross-section (RCS), 521 rain effect, 327-8 ram air, 276
recovery, 302 rectilinear flight, 75-6
Index
refuelling, 295 regulation, 497-8
sideslip response to rudder transfer function, 101
reinforcement, 247
sideslip subsidence mode, 94
relative density, 214, 231 reliability, 527
sightseeing role, 486, 493 single-point mooring, 305, 519
relief valves, 180
size effects, 169, 354-5
remote control systems, 201 remotely piloted airships (RPAs), 428-33 remotely piloted vehicles (RPVs), 429 replenishing, 295-6 response characteristics, 96-102 response transfer functions, 90
on ground handling, 314-15 sizing matrix, 533 skin friction coefficient, 30 skin friction drag, 137
restraint, 301 Reynolds number (Re), 30, 32, 37, 47, 48,
65, 66, 375 rigid airships, 11, 136, 165 rigidity, 181 ring aerofoil, 124, 125 roll oscillation mode, 101
rolling moment, 78 rolling take-off, 335 rotastats, 392, 399-403
rotating hulls, 404-8 rotor systems, 318 rudder angle, 46 rudder control, 323 rudder transfer function, yaw rate response to, 91 ruddervator, 36-8
safety considerations, 482 safety factors, 174-6 Saratoga, 307 satellite navigation systems, 429
SATNAV, 326 search and rescue role, 488
semi-rigid airships, 11, 136, 165, 218, 422-3 sensor platform capability, 482 Sentinel 1000, 279, 283, 285, 286, 288, 293 Sentinel 5000, 35, 280, 283, 285-7, 290-2, 418, 434 servicing, 292 inflight, 290 shape of hull, see hull shape shear collar, 197 Shenandoah, 497
side force, 78
543
Skylab, 487, 493 Skyship, 136, 388
Skyship 500, 15, 29, 37, 207 Skyship 500 G-BIHN, 331 Skyship 600, 15, 30, 65, 66, 168, 323, 326, 481 Skyship 600 G-SKSJ, 328 SLAB, 392, 402 slender body theory, 46-8 slipstream, 136 small perturbations, 84-8 snow, 329-30 solar cells, 442, 458-63 characteristics, 458-60
efficiency, 460-2 types, 462-3 Solar Egg drone, 411 solar energy, 409 solar incidence, angle of, 449-50
solar power, 441-74 available propulsive power, 454 available solar energy, 453-4 collector grid network, 463 components, 458-72 cost considerations, 472 efficiency, 454 electric motors, 463 example calculations, 455-7 projected area, 450-3 storage units, 464 weight considerations, 470-2 solar radiation, 444-8
diffuse, 446-7 direct, 445-6
reflected, 448 sovereignty enforcement role, 489, 496 Spacial MLA-32-B Toluca, 388 SPAS-4, 391
544
Airship Technology
specific fuel consumption (sfc), 129, 349, 351
system engineering, 527 systems, 253, 273-96
speed, 481 response to thrust, 90 response to thrust transfer function, 98
tail drag compared to bare hull drag, 36 tail group, 198-201
spherical shape, 387, 391-2 stability, 73-106, 319 derivatives, 106
control surfaces, 200-1 weight estimation, 249-50 tail lift coefficient, 52
Stability Augmentation System (SAS), 429 stalling characteristics, 122 Standard atmosphere, 229-30 Starship, 419 state equation, 89
tail rigging, 250 tail structure, 249-50 tail surface structural arrangement, 199 take-off, 324, 327, 330, 332, 334-7 take-off power (TOP), 347, 349
static heaviness, 217, 323-5, 338 static lift, 216-17 static lightness, 324-5, 340-1
technology interdependence, 531 technology trends, 528-30 Tedlar, 150
station-keeping, 333 steady flow testing, 65-7 steady-state analytical model, 49 stern propulsion, 377-9
STOL, 392, 394-6 storage tanks, 282, 283 streamline flow, 59, 60 strobe lights, 250 structural categories, 12 structural configuration, 417-26 structural design, 297-8
tee chest, 277, 278 temperature effects, 330, 353 temperature inversions, 329 temperature range, 241 textile materials, 141-53 fracture mechanics approach, 160 textile skin, 11 thermodynamic cycles, 129 thermoplane, 389 thickness ratio, 30, 32 thrust, 127, 135 duct, 127
structures, 165-209
thrust vectoring, 15, 258, 295, 299, 331-3,
fragility, 169-70 general principles and considerations, 166-77 historical, 166 non-rigid airship, 177-209 sub-assemblies, weight percentage analysis of, 245 Sub-Clutter Visibility (SCV) limit, 524
341, 478 thruster system, 136, 316 Titan, 388 Toroidal Balloon Concept, 403 Toroidal Semi-Buoyant Station, 403 toroidal-shaped airships, 392 Torridal ]-rotor airship, 403 trade studies, 521, 526 transfer function, 69 transfer function matrix, 89
stern thrusters, 136
sunshine, 330 Sunship, 409-11, 442-3, 448, 450, 453, 457 super pressure, 177 superheat, 10, 14, 222, 330, 338 surface pressures, 37 surface roughness, 32
transfer lines, 282
transmission systems, 258 trim conditions lateral, 86-7 longitudinal, 84
surge mode, 9]
trim equilibrium, 79, 81, 82
survey role, 487, 493 suspension system, 15, 243 weight estimation, 247-8
trimming, 323 triple-hull airship, 391 turbulence, 55-8, 70
synthetic materials, 142, 143
twin-hull airships, 391
Index
545
tyre scrubbing, 207-9
weight and balance handbook, 270
ultimate acceleration factors, 175 unconventional designs, 385-440
weight estimation, 242-66, 517, 519, 532 and control, 352 ballast, 262-4
undercarriage, 204
cabin, 251-2
unit lift, 6
control system, 258-60
unloading, see loading/unloading
electrical system, 260-2
unmanned airships, 428-33
gondola, 250-3, 267-8
unsteady aerodynamics, 55-7 unsteady flow testing, 67-70
hull, 243-9, 268 propulsion system, 253-8, 350
UPship, 336-7, 423 US Navy airships, 480, 482, 497
suspension system, 247-8 tail group, 249-50
weight percentage analysis of subvalves, 279, 280, 282, 290
assemblies, 245
opening pressures, 279 van Dusen project, 478
Weissinger's approximation, 127 White Dwarf, 413, 414, 486
vector systems, 258, 295
winches, 296
vectored landing, 341-4
wind dimpling, 431
vectored take-off, 336 vectored thrust, 15, 258, 295, 299, 331-3,
wind effects, 326-7, 333-4, 340-4, 356-7 wind speed, 431
341, 478 vertical landing, 344 vertical tunnels, 290 virtual mass, 76, 77 viscous effects, 47 volume, 516
vortex roll-up, 117 vortex wake modelling, 119 VTOL, 392, 394, 396, 400 wake contraction, 118
water ballast dump controls, 288 water recovery, 264, 284
WDL airships, 283 WDL-1 airship JA1005, 325 weather effects, 222, 327-30 weight, 172-3, 235-71, 284, 418 airship operations aspects, 269-71
for take-off and landing, 327 wind tunnel tests, 33, 37, 47, 64-70, 120, 122, 374-5 winged hulls, 397 winged hybrid, 19, 20 wrinkling, 187
yaw rate response to rudder transfer function, 91, 101 yaw subsidence mode, 93 yawing moment, 79 Aalto OS 1OONOU RaoS Wis 7)
Zeppelin New Technology (NT), 136, 419, 424 Zeppy I, 413, 414 Zeppy Il, 413, 414 ZMC-2, 11, 419, 422
antenna, 525 definitions, 235-8 design considerations, 239-42 design criteria, 239
ZPG, 280 ZPG-2, 22, 482 ZPG-2/2W, 283, 290 ZPG-3, 482
empty, 518
ZPG-3W, 193, 198, 282, 418, 434, 482
maximum, 242
ZRS-4, 295
monitoring and control, 267-9
ZRS-5, 295
operations aspects, 269-71 solar power, 470-2 terminology, 235-8 versus volume trend, 528
ZTOL airships, 346
Printed in the United States 30086LVS00002B/11-18
SAN DIEGO PUBLIC LIBRARY
.
:
|
_ PaperbackRe-issue
his book provides a — and indispensable guide to modern “7 desi and operation.
A irships today incorporate advanced technology including composite materials, complex electronic systems and fly-by-light controls. They demand the latest — theoriesin aerodynamics, stability and control and require the use of advanced =
design tools such as numerical finite element structural analysis and computer aided design. This comprehensive and fully illustrated account brings together airship specialists from both universities and industry. After a general introduction, the _ essentials of aerostatics, aerodynamics, stability and control, propulsion, materials and structures are covered. The following chapters consider weight estimates and
control, ground handling and mooring, systems, performance and piloting. The
==
final chapters examine suggestions for improving airship performance, survey
- 2 : unconventional designs, synthesise various design elements, and look at airship
roles and economic considerations vital for the success of the airshipin the auhe Z - - gue Detailedreferences are ae also included.